Daily Papers

The dynamics of a rain forest is extremely complex involving births, deaths and growth of trees with complex interactions between trees, animals, climate, and environment. We consider the patterns of recruits (new trees) and dead trees between rain forest censuses. For a current census we specify regression models for the conditional intensity of recruits and the conditional probabilities of death given the current trees and spatial covariates. We estimate regression parameters using conditional composite likelihood functions that only involve the conditional first order properties of the data. When constructing assumption lean estimators of covariance matrices of parameter estimates we only need mild assumptions of decaying conditional correlations in space while assumptions regarding correlations over time are avoided by exploiting conditional centering of composite likelihood score functions. Time series of point patterns from rain forest censuses are quite short while each point pattern covers a fairly big spatial region. To obtain asymptotic results we therefore use a central limit theorem for the fixed timespan - increasing spatial domain asymptotic setting. This also allows us to handle the challenge of using stochastic covariates constructed from past point patterns. Conveniently, it suffices to impose weak dependence assumptions on the innovations of the space-time process. We investigate the proposed methodology by simulation studies and applications to rain forest data.

This paper is still under revision

Score-based diffusion models have emerged as effective approaches for both conditional and unconditional generation. Still conditional generation is based on either a specific training of a conditional model or classifier guidance, which requires training a noise-dependent classifier, even when a classifier for uncorrupted data is given. We propose a method that, given a pre-trained unconditional score-based generative model, samples from the conditional distribution under arbitrary logical constraints, without requiring additional training. Differently from other zero-shot techniques, that rather aim at generating valid conditional samples, our method is designed for approximating the true conditional distribution. Firstly, we show how to manipulate the learned score in order to sample from an un-normalized distribution conditional on a user-defined constraint. Then, we define a flexible and numerically stable neuro-symbolic framework for encoding soft logical constraints. Combining these two ingredients we obtain a general, but approximate, conditional sampling algorithm. We further developed effective heuristics aimed at improving the approximation. Finally, we show the effectiveness of our approach in approximating conditional distributions for various types of constraints and data: tabular data, images and time series.

Climate change poses complex challenges, with extreme weather events becoming increasingly frequent and difficult to model. Examples include the dynamics of Combined Sewer Systems (CSS). Overburdened CSS during heavy rainfall will overflow untreated wastewater into surface water bodies. Classical approaches to modeling the impact of extreme rainfall events rely on physical simulations, which are particularly challenging to create for large urban infrastructures. Deep Learning (DL) models offer a cost-effective alternative for modeling the complex dynamics of sewer systems. In this study, we present a comprehensive empirical evaluation of several state-of-the-art DL time series models for predicting sewer system dynamics in a large urban infrastructure, utilizing three years of measurement data. We especially investigate the potential of DL models to maintain predictive precision during network outages by comparing global models, which have access to all variables within the sewer system, and local models, which are limited to data from a restricted set of local sensors. Our findings demonstrate that DL models can accurately predict the dynamics of sewer system load, even under network outage conditions. These results suggest that DL models can effectively aid in balancing the load redistribution in CSS, thereby enhancing the sustainability and resilience of urban infrastructures.

8 pages, 4 figures, accepted at 47th German Conference on Artificial Intelligence, Wuerzburg 2024

Anomaly detection (AD) is a fast growing and popular domain among established applications like vision and time series. We observe a rich literature for these applications, but anomaly detection in text is only starting to blossom. Recently, self-supervised methods with self-attention mechanism have been the most popular choice. While recent works have proposed a working ground for building and benchmarking state of the art approaches, we propose two principal contributions in this paper: contextual anomaly contamination and a novel ensemble-based approach. Our method, Textual Anomaly Contamination (TAC), allows to contaminate inlier classes with either independent or contextual anomalies. In the literature, it appears that this distinction is not performed. For finding contextual anomalies, we propose RoSAE, a Robust Subspace Local Recovery Autoencoder Ensemble. All autoencoders of the ensemble present a different latent representation through local manifold learning. Benchmark shows that our approach outperforms recent works on both independent and contextual anomalies, while being more robust. We also provide 8 dataset comparison instead of only relying to Reuters and 20 Newsgroups corpora.

We consider the sparse estimation for stochastic processes with possibly infinite-dimensional nuisance parameters, by using the Dantzig selector which is a sparse estimation method similar to $Z$-estimation. When a consistent estimator for a nuisance parameter is obtained, it is possible to construct an asymptotically normal estimator for the parameter of interest under appropriate conditions. Motivated by this fact, we establish the asymptotic behavior of the Dantzig selector for models of ergodic stochastic processes with high-dimensional parameters of interest and possibly infinite-dimensional nuisance parameters. Moreover, we construct an asymptotically normal estimator by the two step estimation with help of the variable selection through the Dantzig selector and a consistent estimator of the nuisance parameter. Applications to ergodic time series models including integer-valued autoregressive models and ergodic diffusion processes are presented.

Sequential models such as recurrent neural networks or transformer-based models became \textit{de facto} tools for multivariate time series forecasting in a probabilistic fashion, with applications to a wide range of datasets, such as finance, biology, medicine, etc. Despite their adeptness in capturing dependencies, assessing prediction uncertainty, and efficiency in training, challenges emerge in modeling high-dimensional complex distributions and cross-feature dependencies. To tackle these issues, recent works delve into generative modeling by employing diffusion or flow-based models. Notably, the integration of stochastic differential equations or probability flow successfully extends these methods to probabilistic time series imputation and forecasting. However, scalability issues necessitate a computational-friendly framework for large-scale generative model-based predictions. This work proposes a novel approach by blending the computational efficiency of recurrent neural networks with the high-quality probabilistic modeling of the diffusion model, which addresses challenges and advances generative models' application in time series forecasting. Our method relies on the foundation of stochastic interpolants and the extension to a broader conditional generation framework with additional control features, offering insights for future developments in this dynamic field.

We introduce a compositional power transformation, known as an {\alpha}-transformation, to model and forecast a time series of life-table death counts, possibly with zero counts observed at older ages. As a generalisation of the isometric log-ratio transformation (i.e., {\alpha} = 0), the {\alpha} transformation relies on the tuning parameter {\alpha}, which can be determined in a data-driven manner. Using the Australian age-specific period life-table death counts from 1921 to 2020, the {\alpha} transformation can produce more accurate short-term point and interval forecasts than the log-ratio transformation. The improved forecast accuracy of life-table death counts is of great importance to demographers and government planners for estimating survival probabilities and life expectancy and actuaries for determining annuity prices and reserves for various initial ages and maturity terms.

25 pages, 6 tables, 5 figures

Symbolic encoding has been used in multi-operator learning as a way to embed additional information for distinct time-series data. For spatiotemporal systems described by time-dependent partial differential equations, the equation itself provides an additional modality to identify the system. The utilization of symbolic expressions along side time-series samples allows for the development of multimodal predictive neural networks. A key challenge with current approaches is that the symbolic information, i.e. the equations, must be manually preprocessed (simplified, rearranged, etc.) to match and relate to the existing token library, which increases costs and reduces flexibility, especially when dealing with new differential equations. We propose a new token library based on SymPy to encode differential equations as an additional modality for time-series models. The proposed approach incurs minimal cost, is automated, and maintains high prediction accuracy for forecasting tasks. Additionally, we include a Bayesian filtering module that connects the different modalities to refine the learned equation. This improves the accuracy of the learned symbolic representation and the predicted time-series.

The Forward-Forward (FF) algorithm is a recent, purely forward-mode learning method, that updates weights locally and layer-wise and supports supervised as well as unsupervised learning. These features make it ideal for applications such as brain-inspired learning, low-power hardware neural networks, and distributed learning in large models. However, while FF has shown promise on written digit recognition tasks, its performance on natural images and time-series remains a challenge. A key limitation is the need to generate high-quality negative examples for contrastive learning, especially in unsupervised tasks, where versatile solutions are currently lacking. To address this, we introduce the Self-Contrastive Forward-Forward (SCFF) method, inspired by self-supervised contrastive learning. SCFF generates positive and negative examples applicable across different datasets, surpassing existing local forward algorithms for unsupervised classification accuracy on MNIST (MLP: 98.7%), CIFAR-10 (CNN: 80.75%), and STL-10 (CNN: 77.3%). Additionally, SCFF is the first to enable FF training of recurrent neural networks, opening the door to more complex tasks and continuous-time video and text processing.

Attention based models have achieved many remarkable breakthroughs in numerous applications. However, the quadratic complexity of Attention makes the vanilla Attention based models hard to apply to long sequence tasks. Various improved Attention structures are proposed to reduce the computation cost by inducing low rankness and approximating the whole sequence by sub-sequences. The most challenging part of those approaches is maintaining the proper balance between information preservation and computation reduction: the longer sub-sequences used, the better information is preserved, but at the price of introducing more noise and computational costs. In this paper, we propose a smoothed skeleton sketching based Attention structure, coined S$^3$Attention, which significantly improves upon the previous attempts to negotiate this trade-off. S$^3$Attention has two mechanisms to effectively minimize the impact of noise while keeping the linear complexity to the sequence length: a smoothing block to mix information over long sequences and a matrix sketching method that simultaneously selects columns and rows from the input matrix. We verify the effectiveness of S$^3$Attention both theoretically and empirically. Extensive studies over Long Range Arena (LRA) datasets and six time-series forecasting show that S$^3$Attention significantly outperforms both vanilla Attention and other state-of-the-art variants of Attention structures.

Multi-modal large language models (MLLMs) have enabled numerous advances in understanding and reasoning in domains like vision, but we have not yet seen this broad success for time-series. Although prior works on time-series MLLMs have shown promising performance in time-series forecasting, very few works show how an LLM could be used for time-series reasoning in natural language. We propose a novel multi-modal time-series LLM approach that learns generalizable information across various domains with powerful zero-shot performance. First, we train a lightweight time-series encoder on top of an LLM to directly extract time-series information. Then, we fine-tune our model with chain-of-thought augmented time-series tasks to encourage the model to generate reasoning paths. We show that our model learns a latent representation that reflects specific time-series features (e.g. slope, frequency), as well as outperforming GPT-4o on a set of zero-shot reasoning tasks on a variety of domains.

Time Series Foundation Models (TSFMs) have recently garnered attention for their ability to model complex, large-scale time series data across domains such as retail, finance, and transportation. However, their application to sensitive, domain-specific fields like healthcare remains challenging, primarily due to the difficulty of fine-tuning these models for specialized, out-of-domain tasks with scarce publicly available datasets. In this work, we explore the use of Parameter-Efficient Fine-Tuning (PEFT) techniques to address these limitations, focusing on healthcare applications, particularly ICU vitals forecasting for sepsis patients. We introduce and evaluate two selective (BitFit and LayerNorm Tuning) and two additive (VeRA and FourierFT) PEFT techniques on multiple configurations of the Chronos TSFM for forecasting vital signs of sepsis patients. Our comparative analysis demonstrates that some of these PEFT methods outperform LoRA in terms of parameter efficiency and domain adaptation, establishing state-of-the-art (SOTA) results in ICU vital forecasting tasks. Interestingly, FourierFT applied to the Chronos (Tiny) variant surpasses the SOTA model while fine-tuning only 2,400 parameters compared to the 700K parameters of the benchmark.

7 pages. Under review

Addressing the challenges posed by climate change, biodiversity loss, and environmental pollution requires comprehensive monitoring and effective data management strategies that are applicable across various scales in environmental system science. This paper introduces a versatile and transferable digital ecosystem for managing time series data, designed to adhere to the FAIR principles (Findable, Accessible, Interoperable, and Reusable). The system is highly adaptable, cloud-ready, and suitable for deployment in a wide range of settings, from small-scale projects to large-scale monitoring initiatives. The ecosystem comprises three core components: the Sensor Management System (SMS) for detailed metadata registration and management; timeIO, a platform for efficient time series data storage, transfer, and real-time visualization; and the System for Automated Quality Control (SaQC), which ensures data integrity through real-time analysis and quality assurance. The modular architecture, combined with standardized protocols and interfaces, ensures that the ecosystem can be easily transferred and deployed across different environments and institutions. This approach enhances data accessibility for a broad spectrum of stakeholders, including researchers, policymakers, and the public, while fostering collaboration and advancing scientific research in environmental monitoring.

Time position embeddings capture the positional information of time steps, often serving as auxiliary inputs to enhance the predictive capabilities of time series models. However, existing models exhibit limitations in capturing intricate time positional information and effectively utilizing these embeddings. To address these limitations, this paper proposes a novel model called D2Vformer. Unlike typical prediction methods that rely on RNNs or Transformers, this approach can directly handle scenarios where the predicted sequence is not adjacent to the input sequence or where its length dynamically changes. In comparison to conventional methods, D2Vformer undoubtedly saves a significant amount of training resources. In D2Vformer, the Date2Vec module uses the timestamp information and feature sequences to generate time position embeddings. Afterward, D2Vformer introduces a new fusion block that utilizes an attention mechanism to explore the similarity in time positions between the embeddings of the input sequence and the predicted sequence, thereby generating predictions based on this similarity. Through extensive experiments on six datasets, we demonstrate that Date2Vec outperforms other time position embedding methods, and D2Vformer surpasses state-of-the-art methods in both fixed-length and variable-length prediction tasks.

Modelling longitudinal data is an important yet challenging task. These datasets can be high-dimensional, contain non-linear effects and time-varying covariates. Gaussian process (GP) prior-based variational autoencoders (VAEs) have emerged as a promising approach due to their ability to model time-series data. However, they are costly to train and struggle to fully exploit the rich covariates characteristic of longitudinal data, making them difficult for practitioners to use effectively. In this work, we leverage linear mixed models (LMMs) and amortized variational inference to provide conditional priors for VAEs, and propose LMM-VAE, a scalable, interpretable and identifiable model. We highlight theoretical connections between it and GP-based techniques, providing a unified framework for this class of methods. Our proposal performs competitively compared to existing approaches across simulated and real-world datasets.

Deep learning for time-series anomaly detection (TSAD) has gained significant attention over the past decade. Despite the reported improvements in several papers, the practical application of these models remains limited. Recent studies have cast doubt on these models, attributing their results to flawed evaluation techniques. However, the impact of initialization has largely been overlooked. This paper provides a critical analysis of the initialization effects on TSAD model performance. Our extensive experiments reveal that TSAD models are highly sensitive to hyperparameters such as window size, seed number, and normalization. This sensitivity often leads to significant variability in performance, which can be exploited to artificially inflate the reported efficacy of these models. We demonstrate that even minor changes in initialization parameters can result in performance variations that overshadow the claimed improvements from novel model architectures. Our findings highlight the need for rigorous evaluation protocols and transparent reporting of preprocessing steps to ensure the reliability and fairness of anomaly detection methods. This paper calls for a more cautious interpretation of TSAD advancements and encourages the development of more robust and transparent evaluation practices to advance the field and its practical applications.

Tiny Machine Learning (TinyML) enables efficient, lowcost, and privacy preserving machine learning inference directly on microcontroller units (MCUs) connected to sensors. Optimizing models for these constrained environments is crucial. This paper investigates how reducing data acquisition rates affects TinyML models for time series classification, focusing on resource-constrained, battery operated IoT devices. By lowering data sampling frequency, we aim to reduce computational demands RAM usage, energy consumption, latency, and MAC operations by approximately fourfold while maintaining similar classification accuracies. Our experiments with six benchmark datasets (UCIHAR, WISDM, PAMAP2, MHEALTH, MITBIH, and PTB) showed that reducing data acquisition rates significantly cut energy consumption and computational load, with minimal accuracy loss. For example, a 75\% reduction in acquisition rate for MITBIH and PTB datasets led to a 60\% decrease in RAM usage, 75\% reduction in MAC operations, 74\% decrease in latency, and 70\% reduction in energy consumption, without accuracy loss. These results offer valuable insights for deploying efficient TinyML models in constrained environments.

We propose a novel cointegrated autoregressive model for matrix-valued time series, with bi-linear cointegrating vectors corresponding to the rows and columns of the matrix data. Compared to the traditional cointegration analysis, our proposed matrix cointegration model better preserves the inherent structure of the data and enables corresponding interpretations. To estimate the cointegrating vectors as well as other coefficients, we introduce two types of estimators based on least squares and maximum likelihood. We investigate the asymptotic properties of the cointegrated matrix autoregressive model under the existence of trend and establish the asymptotic distributions for the cointegrating vectors, as well as other model parameters. We conduct extensive simulations to demonstrate its superior performance over traditional methods. In addition, we apply our proposed model to Fama-French portfolios and develop a effective pairs trading strategy.

This study presents a comprehensive methodology for modeling and forecasting the historical time series of active fire spots detected by the AQUA_M-T satellite in the Amazon, Brazil. The approach employs a mixed Recurrent Neural Network (RNN) model, combining Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) architectures to predict the monthly accumulations of daily detected active fire spots. Data analysis revealed a consistent seasonality over time, with annual maximum and minimum values tending to repeat at the same periods each year. The primary objective is to verify whether the forecasts capture this inherent seasonality through machine learning techniques. The methodology involved careful data preparation, model configuration, and training using cross-validation with two seeds, ensuring that the data generalizes well to both the test and validation sets for both seeds. The results indicate that the combined LSTM and GRU model delivers excellent forecasting performance, demonstrating its effectiveness in capturing complex temporal patterns and modeling the observed time series. This research significantly contributes to the application of deep learning techniques in environmental monitoring, specifically in forecasting active fire spots. The proposed approach highlights the potential for adaptation to other time series forecasting challenges, opening new opportunities for research and development in machine learning and prediction of natural phenomena. Keywords: Time Series Forecasting; Recurrent Neural Networks; Deep Learning.

16 pages and 24 figures, in Portuguese language

Recently, time series foundation models have shown promising zero-shot forecasting performance on time series from a wide range of domains. However, it remains unclear whether their success stems from a true understanding of temporal dynamics or simply from memorizing the training data. While implicit reasoning in language models has been studied, similar evaluations for time series models have been largely unexplored. This work takes an initial step toward assessing the reasoning abilities of deep time series forecasting models. We find that certain linear, MLP-based, and patch-based Transformer models generalize effectively in systematically orchestrated out-of-distribution scenarios, suggesting underexplored reasoning capabilities beyond simple pattern memorization.

In bipartite causal inference with interference there are two distinct sets of units: those that receive the treatment, termed interventional units, and those on which the outcome is measured, termed outcome units. Which interventional units' treatment can drive which outcome units' outcomes is often depicted in a bipartite network. We study bipartite causal inference with interference from observational data across time and with a changing bipartite network. Under an exposure mapping framework, we define causal effects specific to each outcome unit, representing average contrasts of potential outcomes across time. We establish unconfoundedness of the exposure received by the outcome units based on unconfoundedness assumptions on the interventional units' treatment assignment and the random graph, hence respecting the bipartite structure of the problem. By harvesting the time component of our setting, causal effects are estimable while controlling only for temporal trends and time-varying confounders. Our results hold for binary, continuous, and multivariate exposure mappings. In the case of a binary exposure, we propose three matching algorithms to estimate the causal effect based on matching exposed to unexposed time periods for the same outcome unit, and we show that the bias of the resulting estimators is bounded. We illustrate our approach with an extensive simulation study and an application on the effect of wildfire smoke on transportation by bicycle.

We study noisy calcium imaging data, with a focus on the classification of spike traces. As raw traces obscure the true temporal structure of neuron's activity, we performed a tuned filtering of the calcium concentration using two methods: a biophysical model and a kernel mapping. The former characterizes spike trains related to a particular triggering event, while the latter filters out the signal and refines the selection of the underlying neuronal response. Transitioning from traditional time series analysis to point process theory, the study explores spike-time distance metrics and point pattern prototypes to describe repeated observations. We assume that the analyzed neuron's firing events, i.e. spike occurrences, are temporal point process events. In particular, the study aims to categorize 47 point patterns by depth, assuming the similarity of spike occurrences within specific depth categories. The results highlight the pivotal roles of depth and stimuli in discerning diverse temporal structures of neuron firing events, confirming the point process approach based on prototype analysis is largely useful in the classification of spike traces.

Learning causal relationships between a set of variables is a challenging problem in computer science. Many existing artificial benchmark datasets are based on sampling from causal models and thus contain residual information that the ${R} ^2$-sortability can identify. Here, we present a benchmark for methods in causal learning using time series. The presented dataset is not ${R}^2$-sortable and is based on a real-world scenario of the Krebs cycle that is used in cells to release energy. We provide four scenarios of learning, including short and long time series, and provide guidance so that testing is unified between possible users.

Accurate sports prediction is a crucial skill for professional coaches, which can assist in developing effective training strategies and scientific competition tactics. Traditional methods often use complex mathematical statistical techniques to boost predictability, but this often is limited by dataset scale and has difficulty handling long-term predictions with variable distributions, notably underperforming when predicting point-set-game multi-level matches. To deal with this challenge, this paper proposes TM2, a TCDformer-based Momentum Transfer Model for long-term sports prediction, which encompasses a momentum encoding module and a prediction module based on momentum transfer. TM2 initially encodes momentum in large-scale unstructured time series using the local linear scaling approximation (LLSA) module. Then it decomposes the reconstructed time series with momentum transfer into trend and seasonal components. The final prediction results are derived from the additive combination of a multilayer perceptron (MLP) for predicting trend components and wavelet attention mechanisms for seasonal components. Comprehensive experimental results show that on the 2023 Wimbledon men's tournament datasets, TM2 significantly surpasses existing sports prediction models in terms of performance, reducing MSE by 61.64% and MAE by 63.64%.

Recent works for time-series forecasting more and more leverage the high predictive power of Deep Learning models. With this increase in model complexity, however, comes a lack in understanding of the underlying model decision process, which is problematic for high-stakes decision making. At the same time, simple, interpretable forecasting methods such as Linear Models can still perform very well, sometimes on-par, with Deep Learning approaches. We argue that simple models are good enough most of the time, and forecasting performance can be improved by choosing a Deep Learning method only for certain predictions, increasing the overall interpretability of the forecasting process. In this context, we propose a novel online model selection framework which uses meta-learning to identify these predictions and only rarely uses a non-interpretable, large model. An extensive empirical study on various real-world datasets shows that our selection methodology outperforms state-of-the-art online model selections methods in most cases. We find that almost always choosing a simple Linear Model for forecasting results in competitive performance, suggesting that the need for opaque black-box models in time-series forecasting is smaller than recent works would suggest.

Oxygenators, alarm devices, and footsteps are some of the most common sound sources in a hospital. Detecting them has scientific value for environmental psychology but comes with challenges of its own: namely, privacy preservation and limited labeled data. In this paper, we address these two challenges via a combination of edge computing and cloud computing. For privacy preservation, we have designed an acoustic sensor which computes third-octave spectrograms on the fly instead of recording audio waveforms. For sample-efficient machine learning, we have repurposed a pretrained audio neural network (PANN) via spectral transcoding and label space adaptation. A small-scale study in a neonatological intensive care unit (NICU) confirms that the time series of detected events align with another modality of measurement: i.e., electronic badges for parents and healthcare professionals. Hence, this paper demonstrates the feasibility of polyphonic machine listening in a hospital ward while guaranteeing privacy by design.

We devise a novel formulation and propose the concept of modal participation factors to nonlinear dynamical systems. The original definition of modal participation factors (or simply participation factors) provides a simple yet effective metric. It finds use in theory and practice, quantifying the interplay between states and modes of oscillation in a linear time-invariant (LTI) system. In this paper, with the Koopman operator framework, we present the results of participation factors for nonlinear dynamical systems with an asymptotically stable equilibrium point or limit cycle. We show that participation factors are defined for the entire domain of attraction, beyond the vicinity of an attractor, where the original definition of participation factors for LTI systems is a special case. Finally, we develop a numerical method to estimate participation factors using time series data from the underlying nonlinear dynamical system. The numerical method can be implemented by leveraging a well-established numerical scheme in the Koopman operator framework called dynamic mode decomposition.

Prefetching is a crucial technique employed in traditional databases to enhance interactivity, particularly in the context of data exploitation. Data exploration is a query processing paradigm in which users search for insights buried in the data, often not knowing what exactly they are looking for. Data exploratory tools deal with multiple challenges such as the need for interactivity with no a priori knowledge being present to help with the system tuning. The state-of-the-art prefetchers are specifically designed for navigational workloads only, where the number of possible actions is limited. The prefetchers that work with SQL-based workloads, on the other hand, mainly rely on data logical addresses rather than the data semantics. They fail to predict complex access patterns in cases where the database size is substantial, resulting in an extensive address space, or when there is frequent co-accessing of data. In this paper, we propose SeLeP, a semantic prefetcher that makes prefetching decisions for both types of workloads, based on the encoding of the data values contained inside the accessed blocks. Following the popular path of using machine learning approaches to automatically learn the hidden patterns, we formulate the prefetching task as a time-series forecasting problem and use an encoder-decoder LSTM architecture to learn the data access pattern. Our extensive experiments, across real-life exploratory workloads, demonstrate that SeLeP improves the hit ratio up to 40% and reduces I/O time up to 45% compared to the state-of-the-art, attaining impressive 95% hit ratio and 80% I/O reduction on average.

Modeling spatiotemporal interactions in multivariate time series is key to their effective processing, but challenging because of their irregular and often unknown structure. Statistical properties of the data provide useful biases to model interdependencies and are leveraged by correlation and covariance-based networks as well as by processing pipelines relying on principal component analysis (PCA). However, PCA and its temporal extensions suffer instabilities in the covariance eigenvectors when the corresponding eigenvalues are close to each other, making their application to dynamic and streaming data settings challenging. To address these issues, we exploit the analogy between PCA and graph convolutional filters to introduce the SpatioTemporal coVariance Neural Network (STVNN), a relational learning model that operates on the sample covariance matrix of the time series and leverages joint spatiotemporal convolutions to model the data. To account for the streaming and non-stationary setting, we consider an online update of the parameters and sample covariance matrix. We prove the STVNN is stable to the uncertainties introduced by these online estimations, thus improving over temporal PCA-based methods. Experimental results corroborate our theoretical findings and show that STVNN is competitive for multivariate time series processing, it adapts to changes in the data distribution, and it is orders of magnitude more stable than online temporal PCA.

Joint European Conference on Machine Learning and Knowledge Discovery in Databases (ECML PKDD) 2024

Large language models (LLMs) have received considerable attention recently due to their outstanding comprehension and reasoning capabilities, leading to great progress in many fields. The advancement of LLM techniques also offers promising opportunities to automate many tasks in the telecommunication (telecom) field. After pre-training and fine-tuning, LLMs can perform diverse downstream tasks based on human instructions, paving the way to artificial general intelligence (AGI)-enabled 6G. Given the great potential of LLM technologies, this work aims to provide a comprehensive overview of LLM-enabled telecom networks. In particular, we first present LLM fundamentals, including model architecture, pre-training, fine-tuning, inference and utilization, model evaluation, and telecom deployment. Then, we introduce LLM-enabled key techniques and telecom applications in terms of generation, classification, optimization, and prediction problems. Specifically, the LLM-enabled generation applications include telecom domain knowledge, code, and network configuration generation. After that, the LLM-based classification applications involve network security, text, image, and traffic classification problems. Moreover, multiple LLM-enabled optimization techniques are introduced, such as automated reward function design for reinforcement learning and verbal reinforcement learning. Furthermore, for LLM-aided prediction problems, we discussed time-series prediction models and multi-modality prediction problems for telecom. Finally, we highlight the challenges and identify the future directions of LLM-enabled telecom networks.

Multivariate time series data suffer from the problem of missing values, which hinders the application of many analytical methods. To achieve the accurate imputation of these missing values, exploiting inter-correlation by employing the relationships between sequences (i.e., a network) is as important as the use of temporal dependency, since a sequence normally correlates with other sequences. Moreover, exploiting an adequate network depending on time is also necessary since the network varies over time. However, in real-world scenarios, we normally know neither the network structure nor when the network changes beforehand. Here, we propose a missing value imputation method for multivariate time series, namely MissNet, that is designed to exploit temporal dependency with a state-space model and inter-correlation by switching sparse networks. The network encodes conditional independence between features, which helps us understand the important relationships for imputation visually. Our algorithm, which scales linearly with reference to the length of the data, alternatively infers networks and fills in missing values using the networks while discovering the switching of the networks. Extensive experiments demonstrate that MissNet outperforms the state-of-the-art algorithms for multivariate time series imputation and provides interpretable results.

Pushing the frontiers of time-series information processing in the ever-growing domain of edge devices with stringent resources has been impeded by the systems' ability to process information and learn locally on the device. Local processing and learning of time-series information typically demand intensive computations and massive storage as the process involves retrieving information and tuning hundreds of parameters back in time. In this work, we developed a memristor-based echo state network accelerator that features efficient temporal data processing and in-situ online learning. The proposed design is benchmarked using various datasets involving real-world tasks, such as forecasting the load energy consumption and weather conditions. The experimental results illustrate that the hardware model experiences a marginal degradation in performance as compared to the software counterpart. This is mainly attributed to the limited precision and dynamic range of network parameters when emulated using memristor devices. The proposed system is evaluated for lifespan, robustness, and energy-delay product. It is observed that the system demonstrates reasonable robustness for device failure below 10%, which may occur due to stuck-at faults. Furthermore, 247X reduction in energy consumption is achieved when compared to a custom CMOS digital design implemented at the same technology node.

Time series are ubiquitous and occur naturally in a variety of applications -- from data recorded by sensors in manufacturing processes, over financial data streams to climate data. Different tasks arise, such as regression, classification or segmentation of the time series. However, to reliably solve these challenges, it is important to filter out abnormal observations that deviate from the usual behavior of the time series. While many anomaly detection methods exist for independent data and stationary time series, these methods are not applicable to non-stationary time series. To allow for non-stationarity in the data, while simultaneously detecting anomalies, we propose OML-AD, a novel approach for anomaly detection (AD) based on online machine learning (OML). We provide an implementation of OML-AD within the Python library River and show that it outperforms state-of-the-art baseline methods in terms of accuracy and computational efficiency.

14 pages, 4 figures, 4 tables

Satellite image time series (SITS) data provides continuous observations over time, allowing for the tracking of vegetation changes and growth patterns throughout the seasons and years. Numerous deep learning (DL) approaches using SITS for crop classification have emerged recently, with the latest approaches adopting Transformer for SITS classification. However, the quadratic complexity of self-attention in Transformer poses challenges for classifying long time series. While the cutting-edge Mamba architecture has demonstrated strength in various domains, including remote sensing image interpretation, its capacity to learn temporal representations in SITS data remains unexplored. Moreover, the existing SITS classification methods often depend solely on crop labels as supervision signals, which fails to fully exploit the temporal information. In this paper, we proposed a Satellite Image Time Series Mamba (SITSMamba) method for crop classification based on remote sensing time series data. The proposed SITSMamba contains a spatial encoder based on Convolutional Neural Networks (CNN) and a Mamba-based temporal encoder. To exploit richer temporal information from SITS, we design two branches of decoder used for different tasks. The first branch is a crop Classification Branch (CBranch), which includes a ConvBlock to decode the feature to a crop map. The second branch is a SITS Reconstruction Branch that uses a Linear layer to transform the encoded feature to predict the original input values. Furthermore, we design a Positional Weight (PW) applied to the RBranch to help the model learn rich latent knowledge from SITS. We also design two weighting factors to control the balance of the two branches during training. The code of SITSMamba is available at: https://github.com/XiaoleiQinn/SITSMamba.

Multivariate time series classification is an important task with widespread domains of applications. Recently, deep neural networks (DNN) have achieved state-of-the-art performance in time series classification. However, they often require large expert-labeled training datasets which can be infeasible in practice. In few-shot settings, i.e. only a limited number of samples per class are available in training data, DNNs show a significant drop in testing accuracy and poor generalization ability. In this paper, we propose to address these problems from an optimization and a loss function perspective. Specifically, we propose a new learning framework named COSCO consisting of a sharpness-aware minimization (SAM) optimization and a Prototypical loss function to improve the generalization ability of DNN for multivariate time series classification problems under few-shot setting. Our experiments demonstrate our proposed method outperforms the existing baseline methods. Our source code is available at: https://github.com/JRB9/COSCO.

5 pages, 5 figures, CIKM '24 Short Paper Track

Transformers for time series forecasting mainly model time series from limited or fixed scales, making it challenging to capture different characteristics spanning various scales. We propose Pathformer, a multi-scale Transformer with adaptive pathways. It integrates both temporal resolution and temporal distance for multi-scale modeling. Multi-scale division divides the time series into different temporal resolutions using patches of various sizes. Based on the division of each scale, dual attention is performed over these patches to capture global correlations and local details as temporal dependencies. We further enrich the multi-scale Transformer with adaptive pathways, which adaptively adjust the multi-scale modeling process based on the varying temporal dynamics of the input, improving the accuracy and generalization of Pathformer. Extensive experiments on eleven real-world datasets demonstrate that Pathformer not only achieves state-of-the-art performance by surpassing all current models but also exhibits stronger generalization abilities under various transfer scenarios. The code is made available at https://github.com/decisionintelligence/pathformer.

Accepted by the 12th International Conference on Learning Representations (ICLR 2024)

Temporal exponential random graph models (TERGM) are powerful statistical models that can be used to infer the temporal pattern of edge formation and elimination in complex networks (e.g., social networks). TERGMs can also be used in a generative capacity to predict longitudinal time series data in these evolving graphs. However, parameter estimation within this framework fails to capture many real-world properties of social networks, including: triadic relationships, small world characteristics, and social learning theories which could be used to constrain the probabilistic estimation of dyadic covariates. Here, we propose triadic temporal exponential random graph models (TTERGM) to fill this void, which includes these hierarchical network relationships within the graph model. We represent social network learning theory as an additional probability distribution that optimizes Markov chains in the graph vector space. The new parameters are then approximated via Monte Carlo maximum likelihood estimation. We show that our TTERGM model achieves improved fidelity and more accurate predictions compared to several benchmark methods on GitHub network data.

In time-series classification, understanding model decisions is crucial for their application in high-stakes domains such as healthcare and finance. Counterfactual explanations, which provide insights by presenting alternative inputs that change model predictions, offer a promising solution. However, existing methods for generating counterfactual explanations for time-series data often struggle with balancing key objectives like proximity, sparsity, and validity. In this paper, we introduce TX-Gen, a novel algorithm for generating counterfactual explanations based on the Non-dominated Sorting Genetic Algorithm II (NSGA-II). TX-Gen leverages evolutionary multi-objective optimization to find a diverse set of counterfactuals that are both sparse and valid, while maintaining minimal dissimilarity to the original time series. By incorporating a flexible reference-guided mechanism, our method improves the plausibility and interpretability of the counterfactuals without relying on predefined assumptions. Extensive experiments on benchmark datasets demonstrate that TX-Gen outperforms existing methods in generating high-quality counterfactuals, making time-series models more transparent and interpretable.

Preprint, under review

Accurate demand forecasting is crucial for optimizing supply chain management. Traditional methods often fail to capture complex patterns from seasonal variability and special events. Despite advancements in deep learning, interpretable forecasting models remain a challenge. To address this, we introduce the Multi-Channel Data Fusion Network (MCDFN), a hybrid architecture that integrates Convolutional Neural Networks (CNN), Long Short-Term Memory networks (LSTM), and Gated Recurrent Units (GRU) to enhance predictive performance by extracting spatial and temporal features from time series data. Our comparative benchmarking demonstrates that MCDFN outperforms seven other deep-learning models, achieving superior metrics: MSE (23.5738), RMSE (4.8553), MAE (3.9991), and MAPE (20.1575%). Additionally, MCDFN's predictions were statistically indistinguishable from actual values, confirmed by a paired t-test with a 5% p-value and a 10-fold cross-validated statistical paired t-test. We apply explainable AI techniques like ShapTime and Permutation Feature Importance to enhance interpretability. This research advances demand forecasting methodologies and offers practical guidelines for integrating MCDFN into supply chain systems, highlighting future research directions for scalability and user-friendly deployment.

Archaeological sites are the physical remains of past human activity and one of the main sources of information about past societies and cultures. However, they are also the target of malevolent human actions, especially in countries having experienced inner turmoil and conflicts. Because monitoring these sites from space is a key step towards their preservation, we introduce the DAFA Looted Sites dataset, \datasetname, a labeled multi-temporal remote sensing dataset containing 55,480 images acquired monthly over 8 years across 675 Afghan archaeological sites, including 135 sites looted during the acquisition period. \datasetname~is particularly challenging because of the limited number of training samples, the class imbalance, the weak binary annotations only available at the level of the time series, and the subtlety of relevant changes coupled with important irrelevant ones over a long time period. It is also an interesting playground to assess the performance of satellite image time series (SITS) classification methods on a real and important use case. We evaluate a large set of baselines, outline the substantial benefits of using foundation models and show the additional boost that can be provided by using complete time series instead of using a single image.

Performance testing aims at uncovering efficiency issues of software systems. In order to be both effective and practical, the design of a performance test must achieve a reasonable trade-off between result quality and testing time. This becomes particularly challenging in Java context, where the software undergoes a warm-up phase of execution, due to just-in-time compilation. During this phase, performance measurements are subject to severe fluctuations, which may adversely affect quality of performance test results. However, these approaches often provide suboptimal estimates of the warm-up phase, resulting in either insufficient or excessive warm-up iterations, which may degrade result quality or increase testing time. There is still a lack of consensus on how to properly address this problem. Here, we propose and study an AI-based framework to dynamically halt warm-up iterations at runtime. Specifically, our framework leverages recent advances in AI for Time Series Classification (TSC) to predict the end of the warm-up phase during test execution. We conduct experiments by training three different TSC models on half a million of measurement segments obtained from JMH microbenchmark executions. We find that our framework significantly improves the accuracy of the warm-up estimates provided by state-of-practice and state-of-the-art methods. This higher estimation accuracy results in a net improvement in either result quality or testing time for up to +35.3% of the microbenchmarks. Our study highlights that integrating AI to dynamically estimate the end of the warm-up phase can enhance the cost-effectiveness of Java performance testing.

Accepted for publication in The 39th IEEE/ACM International Conference on Automated Software Engineering (ASE '24)

As global climate change intensifies, accurate weather forecasting is increasingly crucial for sectors such as agriculture, energy management, and environmental protection. Traditional methods, which rely on physical and statistical models, often struggle with complex, nonlinear, and time-varying data, underscoring the need for more advanced techniques. This study explores a hybrid CNN-LSTM model to enhance temperature forecasting accuracy for the Delhi region, using historical meteorological data from 1996 to 2017. We employed both direct and indirect methods, including comprehensive data preprocessing and exploratory analysis, to construct and train our model. The CNN component effectively extracts spatial features, while the LSTM captures temporal dependencies, leading to improved prediction accuracy. Experimental results indicate that the CNN-LSTM model significantly outperforms traditional forecasting methods in terms of both accuracy and stability, with a mean square error (MSE) of 3.26217 and a root mean square error (RMSE) of 1.80615. The hybrid model demonstrates its potential as a robust tool for temperature prediction, offering valuable insights for meteorological forecasting and related fields. Future research should focus on optimizing model architecture, exploring additional feature extraction techniques, and addressing challenges such as overfitting and computational complexity. This approach not only advances temperature forecasting but also provides a foundation for applying deep learning to other time series forecasting tasks.

Statistical models are an essential tool to model, forecast and understand the hydrological processes in watersheds. In particular, the understanding of time lags associated with the delay between rainfall occurrence and subsequent changes in streamflow is of high practical importance. Since water can take a variety of flow paths to generate streamflow, a series of distinct runoff pulses may combine to create the observed streamflow time series. Current state-of-the-art models are not able to sufficiently confront the problem complexity with interpretable parametrization, thus preventing novel insights about the dynamics of distinct flow paths from being formed. The proposed Gaussian Sliding Windows Regression Model targets this problem by combining the concept of multiple windows sliding along the time axis with multiple linear regression. The window kernels, which indicate the weights applied to different time lags, are implemented via Gaussian-shaped kernels. As a result, straightforward process inference can be achieved since each window can represent one flow path. Experiments on simulated and real-world scenarios underline that the proposed model achieves accurate parameter estimates and competitive predictive performance, while fostering explainable and interpretable hydrological modeling.

We propose tensor time series imputation when the missing pattern in the tensor data can be general, as long as any two data positions along a tensor fibre are both observed for enough time points. The method is based on a tensor time series factor model with Tucker decomposition of the common component. One distinguished feature of the tensor time series factor model used is that there can be weak factors in the factor loadings matrix for each mode. This reflects reality better when real data can have weak factors which drive only groups of observed variables, for instance, a sector factor in financial market driving only stocks in a particular sector. Using the data with missing entries, asymptotic normality is derived for rows of estimated factor loadings, while consistent covariance matrix estimation enables us to carry out inferences. As a first in the literature, we also propose a ratio-based estimator for the rank of the core tensor under general missing patterns. Rates of convergence are spelt out for the imputations from the estimated tensor factor models. Simulation results show that our imputation procedure works well, with asymptotic normality and corresponding inferences also demonstrated. Re-imputation performances are also gauged when we demonstrate that using slightly larger rank then estimated gives superior re-imputation performances. A Fama-French portfolio example with matrix returns and an OECD data example with matrix of Economic indicators are presented and analyzed, showing the efficacy of our imputation approach compared to direct vector imputation.

Computer network anomaly detection and log analysis, as an important topic in the field of network security, has been a key task to ensure network security and system reliability. First, existing network anomaly detection and log analysis methods are often challenged by high-dimensional data and complex network topologies, resulting in unstable performance and high false-positive rates. In addition, traditional methods are usually difficult to handle time-series data, which is crucial for anomaly detection and log analysis. Therefore, we need a more efficient and accurate method to cope with these problems. To compensate for the shortcomings of current methods, we propose an innovative fusion model that integrates Isolation Forest, GAN (Generative Adversarial Network), and Transformer with each other, and each of them plays a unique role. Isolation Forest is used to quickly identify anomalous data points, and GAN is used to generate synthetic data with the real data distribution characteristics to augment the training dataset, while the Transformer is used for modeling and context extraction on time series data. The synergy of these three components makes our model more accurate and robust in anomaly detection and log analysis tasks. We validate the effectiveness of this fusion model in an extensive experimental evaluation. Experimental results show that our model significantly improves the accuracy of anomaly detection while reducing the false alarm rate, which helps to detect potential network problems in advance. The model also performs well in the log analysis task and is able to quickly identify anomalous behaviors, which helps to improve the stability of the system. The significance of this study is that it introduces advanced deep learning techniques, which work anomaly detection and log analysis.

The Matrix Profile (MP), a versatile tool for time series data mining, has been shown effective in time series anomaly detection (TSAD). This paper delves into the problem of anomaly detection in multidimensional time series, a common occurrence in real-world applications. For instance, in a manufacturing factory, multiple sensors installed across the site collect time-varying data for analysis. The Matrix Profile, named for its role in profiling the matrix storing pairwise distance between subsequences of univariate time series, becomes complex in multidimensional scenarios. If the input univariate time series has n subsequences, the pairwise distance matrix is a n x n matrix. In a multidimensional time series with d dimensions, the pairwise distance information must be stored in a n x n x d tensor. In this paper, we first analyze different strategies for condensing this tensor into a profile vector. We then investigate the potential of extending the MP to efficiently find k-nearest neighbors for anomaly detection. Finally, we benchmark the multidimensional MP against 19 baseline methods on 119 multidimensional TSAD datasets. The experiments covers three learning setups: unsupervised, supervised, and semi-supervised. MP is the only method that consistently delivers high performance across all setups.

Deep learning, a branch of artificial intelligence, is a data-driven method that uses multiple layers of interconnected units (neurons) to learn intricate patterns and representations directly from raw input data. Empowered by this learning capability, it has become a powerful tool for solving complex problems and is the core driver of many groundbreaking technologies and innovations. Building a deep learning model is challenging due to the algorithm's complexity and the dynamic nature of real-world problems. Several studies have reviewed deep learning concepts and applications. However, the studies mostly focused on the types of deep learning models and convolutional neural network architectures, offering limited coverage of the state-of-the-art deep learning models and their applications in solving complex problems across different domains. Therefore, motivated by the limitations, this study aims to comprehensively review the state-of-the-art deep learning models in computer vision, natural language processing, time series analysis and pervasive computing. We highlight the key features of the models and their effectiveness in solving the problems within each domain. Furthermore, this study presents the fundamentals of deep learning, various deep learning model types and prominent convolutional neural network architectures. Finally, challenges and future directions in deep learning research are discussed to offer a broader perspective for future researchers.

Submitted to Applied Soft Computing

Accurate imputation is essential for the reliability and success of downstream tasks. Recently, diffusion models have attracted great attention in this field. However, these models neglect the latent distribution in a lower-dimensional space derived from the observed data, which limits the generative capacity of the diffusion model. Additionally, dealing with the original missing data without labels becomes particularly problematic. To address these issues, we propose the Latent Space Score-Based Diffusion Model (LSSDM) for probabilistic multivariate time series imputation. Observed values are projected onto low-dimensional latent space and coarse values of the missing data are reconstructed without knowing their ground truth values by this unsupervised learning approach. Finally, the reconstructed values are fed into a conditional diffusion model to obtain the precise imputed values of the time series. In this way, LSSDM not only possesses the power to identify the latent distribution but also seamlessly integrates the diffusion model to obtain the high-fidelity imputed values and assess the uncertainty of the dataset. Experimental results demonstrate that LSSDM achieves superior imputation performance while also providing a better explanation and uncertainty analysis of the imputation mechanism. The website of the code is \textit{https://github.com/gorgen2020/LSSDM\_imputation}.

Continuous-time series is essential for different modern application areas, e.g. healthcare, automobile, energy, finance, Internet of things (IoT) and other related areas. Different application needs to process as well as analyse a massive amount of data in time series structure in order to determine the data-driven result, for example, financial trend prediction, potential probability of the occurrence of a particular event occurrence identification, patient health record processing and so many more. However, modeling real-time data using a continuous-time series is challenging since the dynamical systems behind the data could be a differential equation. Several research works have tried to solve the challenges of modelling the continuous-time series using different neural network models and approaches for data processing and learning. The existing deep learning models are not free from challenges and limitations due to diversity among different attributes, behaviour, duration of steps, energy, and data sampling rate. This paper has described the general problem domain of time series and reviewed the challenges of modelling the continuous time series. We have presented a comparative analysis of recent developments in deep learning models and their contribution to solving different difficulties of modelling the continuous time series. We have also identified the limitations of the existing neural network model and open issues. The main goal of this review is to understand the recent trend of neural network models used in a different real-world application with continuous-time data.

Event detection in time series is a challenging task due to the prevalence of imbalanced datasets, rare events, and time interval-defined events. Traditional supervised deep learning methods primarily employ binary classification, where each time step is assigned a binary label indicating the presence or absence of an event. However, these methods struggle to handle these specific scenarios effectively. To address these limitations, we propose a novel supervised regression-based deep learning approach that offers several advantages over classification-based methods. Our approach, with a limited number of parameters, can effectively handle various types of events within a unified framework, including rare events and imbalanced datasets. We provide theoretical justifications for its universality and precision and demonstrate its superior performance across diverse domains, particularly for rare events and imbalanced datasets.

Community-based fact-checking is a promising approach to verify social media content and correct misleading posts at scale. Yet, causal evidence regarding its effectiveness in reducing the spread of misinformation on social media is missing. Here, we performed a large-scale empirical study to analyze whether community notes reduce the spread of misleading posts on X. Using a Difference-in-Differences design and repost time series data for N=237,677 (community fact-checked) cascades that had been reposted more than 431 million times, we found that exposing users to community notes reduced the spread of misleading posts by, on average, 62.0%. Furthermore, community notes increased the odds that users delete their misleading posts by 103.4%. However, our findings also suggest that community notes might be too slow to intervene in the early (and most viral) stage of the diffusion. Our work offers important implications to enhance the effectiveness of community-based fact-checking approaches on social media.

Gross domestic product (GDP) nowcasting is crucial for policy-making as GDP growth is a key indicator of economic conditions. Dynamic factor models (DFMs) have been widely adopted by government agencies for GDP nowcasting due to their ability to handle irregular or missing macroeconomic indicators and their interpretability. However, DFMs face two main challenges: i) the lack of capturing economic uncertainties such as sudden recessions or booms, and ii) the limitation of capturing irregular dynamics from mixed-frequency data. To address these challenges, we introduce NCDENow, a novel GDP nowcasting framework that integrates neural controlled differential equations (NCDEs) with DFMs. This integration effectively handles the dynamics of irregular time series. NCDENow consists of 3 main modules: i) factor extraction leveraging DFM, ii) dynamic modeling using NCDE, and iii) GDP growth prediction through regression. We evaluate NCDENow against 6 baselines on 2 real-world GDP datasets from South Korea and the United Kingdom, demonstrating its enhanced predictive capability. Our empirical results favor our method, highlighting the significant potential of integrating NCDE into nowcasting models. Our code and dataset are available at https://github.com/sklim84/NCDENow_CIKM2024.

Accepted at CIKM 2024. Seonkyu Lim and Jeongwhan Choi are co-first authors with equal contributions

The selection of algorithms is a crucial step in designing AI services for real-world time series classification use cases. Traditional methods such as neural architecture search, automated machine learning, combined algorithm selection, and hyperparameter optimizations are effective but require considerable computational resources and necessitate access to all data points to run their optimizations. In this work, we introduce a novel data fingerprint that describes any time series classification dataset in a privacy-preserving manner and provides insight into the algorithm selection problem without requiring training on the (unseen) dataset. By decomposing the multi-target regression problem, only our data fingerprints are used to estimate algorithm performance and uncertainty in a scalable and adaptable manner. Our approach is evaluated on the 112 University of California riverside benchmark datasets, demonstrating its effectiveness in predicting the performance of 35 state-of-the-art algorithms and providing valuable insights for effective algorithm selection in time series classification service systems, improving a naive baseline by 7.32% on average in estimating the mean performance and 15.81% in estimating the uncertainty.

Hawaii International Conference on System Sciences (HICSS-58) 2025

Subspace representation is a fundamental technique in various fields of machine learning. Analyzing a geometrical relationship among multiple subspaces is essential for understanding subspace series' temporal and/or spatial dynamics. This paper proposes the second-order difference subspace, a higher-order extension of the first-order difference subspace between two subspaces that can analyze the geometrical difference between them. As a preliminary for that, we extend the definition of the first-order difference subspace to the more general setting that two subspaces with different dimensions have an intersection. We then define the second-order difference subspace by combining the concept of first-order difference subspace and principal component subspace (Karcher mean) between two subspaces, motivated by the second-order central difference method. We can understand that the first/second-order difference subspaces correspond to the velocity and acceleration of subspace dynamics from the viewpoint of a geodesic on a Grassmann manifold. We demonstrate the validity and naturalness of our second-order difference subspace by showing numerical results on two applications: temporal shape analysis of a 3D object and time series analysis of a biometric signal.

Long-short range time series forecasting is essential for predicting future trends and patterns over extended periods. While deep learning models such as Transformers have made significant strides in advancing time series forecasting, they often encounter difficulties in capturing long-term dependencies and effectively managing sparse semantic features. The state-space model, Mamba, addresses these issues through its adept handling of selective input and parallel computing, striking a balance between computational efficiency and prediction accuracy. This article examines the advantages and disadvantages of both Mamba and Transformer models, and introduces a combined approach, MAT, which leverages the strengths of each model to capture unique long-short range dependencies and inherent evolutionary patterns in multivariate time series. Specifically, MAT harnesses the long-range dependency capabilities of Mamba and the short-range characteristics of Transformers. Experimental results on benchmark weather datasets demonstrate that MAT outperforms existing comparable methods in terms of prediction accuracy, scalability, and memory efficiency.

6 pages, 4 figures, to be presented at the 5th International Conference on Electrical, Communication and Computer Engineering (ICECCE)

Time series generation is a crucial research topic in the area of decision-making systems, which can be particularly important in domains like autonomous driving, healthcare, and, notably, robotics. Recent approaches focus on learning in the data space to model time series information. However, the data space often contains limited observations and noisy features. In this paper, we propose TimeLDM, a novel latent diffusion model for high-quality time series generation. TimeLDM is composed of a variational autoencoder that encodes time series into an informative and smoothed latent content and a latent diffusion model operating in the latent space to generate latent information. We evaluate the ability of our method to generate synthetic time series with simulated and real-world datasets and benchmark the performance against existing state-of-the-art methods. Qualitatively and quantitatively, we find that the proposed TimeLDM persistently delivers high-quality generated time series. For example, TimeLDM achieves new state-of-the-art results on the simulated benchmarks and an average improvement of 55% in Discriminative score with all benchmarks. Further studies demonstrate that our method yields more robust outcomes across various lengths of time series data generation. Especially, for the Context-FID score and Discriminative score, TimeLDM realizes significant improvements of 80% and 50%, respectively. The code will be released after publication.

Daily activity monitoring systems used in households provide vital information for health status, particularly with aging residents. Multiple approaches have been introduced to achieve such goals, typically obtrusive and non-obtrusive. Amongst the obtrusive approaches are the wearable devices, and among the non-obtrusive approaches are the movement detection systems, including motion sensors and thermal sensor arrays (TSAs). TSA systems are advantageous when preserving a person's privacy and picking his precise spatial location. In this study, human daily living activities were monitored day and night, constructing the corresponding activity time series and spatial probability distribution and employing a TSA system. The monitored activities are classified into two categories: sleeping and daily activity. Results showed the possibility of distinguishing between classes regardless of day and night. The obtained sleep activity duration was compared with previous research using the same raw data. Results showed that the duration of sleep activity, on average, was 9 hours/day, and daily life activity was 7 hours/day. The person's spatial probability distribution was determined using the bivariate distribution for the monitored location. In conclusion, the results showed that sleeping activity was dominant. Our study showed that TSAs were the optimum choice when monitoring human activity. Our proposed approach tackled limitations encountered by previous human activity monitoring systems, such as preserving human privacy while knowing his precise spatial location.

Wildfires have significant impacts on global vegetation, wildlife, and humans. They destroy plant communities and wildlife habitats and contribute to increased emissions of carbon dioxide, nitrogen oxides, methane, and other pollutants. The prediction of wildfires relies on various independent variables combined with regression or machine learning methods. In this technical review, we describe the options for independent variables, data processing techniques, models, independent variables collinearity and importance estimation methods, and model performance evaluation metrics. First, we divide the independent variables into 4 aspects, including climate and meteorology conditions, socio-economical factors, terrain and hydrological features, and wildfire historical records. Second, preprocessing methods are described for different magnitudes, different spatial-temporal resolutions, and different formats of data. Third, the collinearity and importance evaluation methods of independent variables are also considered. Fourth, we discuss the application of statistical models, traditional machine learning models, and deep learning models in wildfire risk prediction. In this subsection, compared with other reviews, this manuscript particularly discusses the evaluation metrics and recent advancements in deep learning methods. Lastly, addressing the limitations of current research, this paper emphasizes the need for more effective deep learning time series forecasting algorithms, the utilization of three-dimensional data including ground and trunk fuel, extraction of more accurate historical fire point data, and improved model evaluation metrics.

To make sense of massive data, we often fit simplified models and then interpret the parameters; for example, we cluster the text embeddings and then interpret the mean parameters of each cluster. However, these parameters are often high-dimensional and hard to interpret. To make model parameters directly interpretable, we introduce a family of statistical models -- including clustering, time series, and classification models -- parameterized by natural language predicates. For example, a cluster of text about COVID could be parameterized by the predicate "discusses COVID". To learn these statistical models effectively, we develop a model-agnostic algorithm that optimizes continuous relaxations of predicate parameters with gradient descent and discretizes them by prompting language models (LMs). Finally, we apply our framework to a wide range of problems: taxonomizing user chat dialogues, characterizing how they evolve across time, finding categories where one language model is better than the other, clustering math problems based on subareas, and explaining visual features in memorable images. Our framework is highly versatile, applicable to both textual and visual domains, can be easily steered to focus on specific properties (e.g. subareas), and explains sophisticated concepts that classical methods (e.g. n-gram analysis) struggle to produce.

We introduce VistaFormer, a lightweight Transformer-based model architecture for the semantic segmentation of remote-sensing images. This model uses a multi-scale Transformer-based encoder with a lightweight decoder that aggregates global and local attention captured in the encoder blocks. VistaFormer uses position-free self-attention layers which simplifies the model architecture and removes the need to interpolate temporal and spatial codes, which can reduce model performance when training and testing image resolutions differ. We investigate simple techniques for filtering noisy input signals like clouds and demonstrate that improved model scalability can be achieved by substituting Multi-Head Self-Attention (MHSA) with Neighbourhood Attention (NA). Experiments on the PASTIS and MTLCC crop-type segmentation benchmarks show that VistaFormer achieves better performance than comparable models and requires only 8% of the floating point operations using MHSA and 11% using NA while also using fewer trainable parameters. VistaFormer with MHSA improves on state-of-the-art mIoU scores by 0.1% on the PASTIS benchmark and 3% on the MTLCC benchmark while VistaFormer with NA improves on the MTLCC benchmark by 3.7%.

Unsupervised domain adaptation (UDA) of time series aims to teach models to identify consistent patterns across various temporal scenarios, disregarding domain-specific differences, which can maintain their predictive accuracy and effectively adapt to new domains. However, existing UDA methods struggle to adequately extract and align both global and local features in time series data. To address this issue, we propose the Local-Global Representation Alignment framework (LogoRA), which employs a two-branch encoder, comprising a multi-scale convolutional branch and a patching transformer branch. The encoder enables the extraction of both local and global representations from time series. A fusion module is then introduced to integrate these representations, enhancing domain-invariant feature alignment from multi-scale perspectives. To achieve effective alignment, LogoRA employs strategies like invariant feature learning on the source domain, utilizing triplet loss for fine alignment and dynamic time warping-based feature alignment. Additionally, it reduces source-target domain gaps through adversarial training and per-class prototype alignment. Our evaluations on four time-series datasets demonstrate that LogoRA outperforms strong baselines by up to $12.52\%$, showcasing its superiority in time series UDA tasks.

Accepted by IEEE Transactions on Knowledge and Data Engineering

In deep learning, the load data with non-temporal factors are difficult to process by sequence models. This problem results in insufficient precision of the prediction. Therefore, a short-term load forecasting method based on convolutional neural network (CNN), self-attention encoder-decoder network (SAEDN) and residual-refinement (Res) is proposed. In this method, feature extraction module is composed of a two-dimensional convolutional neural network, which is used to mine the local correlation between data and obtain high-dimensional data features. The initial load fore-casting module consists of a self-attention encoder-decoder network and a feedforward neural network (FFN). The module utilizes self-attention mechanisms to encode high-dimensional features. This operation can obtain the global correlation between data. Therefore, the model is able to retain important information based on the coupling relationship between the data in data mixed with non-time series factors. Then, self-attention decoding is per-formed and the feedforward neural network is used to regression initial load. This paper introduces the residual mechanism to build the load optimization module. The module generates residual load values to optimize the initial load. The simulation results show that the proposed load forecasting method has advantages in terms of prediction accuracy and prediction stability.

in Chinese language, Accepted by Electric Power Automation Equipment

Aiming at the problem that delay time is difficult to determine and prediction accuracy is low in building prediction model of SCR system, a dynamic modeling scheme based on a hybrid of multiple data-driven algorithms was proposed. First, processed abnormal values and normalized the data. To improve the relevance of the input data, used MIC to estimate delay time and reconstructed production data. Then used combined feature selection method to determine input variables. To further mine data information, VMD was used to decompose input time series. Finally, established NOx emission prediction model combining ELM and EC model. Experimental results based on actual historical operating data show that the MAPE of predicted results is 2.61%. Model sensitivity analysis shows that besides the amount of ammonia injection, the inlet oxygen concentration and the flue gas temperature have a significant impact on NOx emission, which should be considered in SCR process control and optimization.

in Chinese language, Accepted by Proceedings of the CSEE

Functional data analysis (FDA) and ensemble learning can be powerful tools for analyzing complex environmental time series. Recent literature has highlighted the key role of diversity in enhancing accuracy and reducing variance in ensemble methods.This paper introduces Randomized Spline Trees (RST), a novel algorithm that bridges these two approaches by incorporating randomized functional representations into the Random Forest framework. RST generates diverse functional representations of input data using randomized B-spline parameters, creating an ensemble of decision trees trained on these varied representations. We provide a theoretical analysis of how this functional diversity contributes to reducing generalization error and present empirical evaluations on six environmental time series classification tasks from the UCR Time Series Archive. Results show that RST variants outperform standard Random Forests and Gradient Boosting on most datasets, improving classification accuracy by up to 14\%. The success of RST demonstrates the potential of adaptive functional representations in capturing complex temporal patterns in environmental data. This work contributes to the growing field of machine learning techniques focused on functional data and opens new avenues for research in environmental time series analysis.

The use of in-situ digital sensors for water quality monitoring is becoming increasingly common worldwide. While these sensors provide near real-time data for science, the data are prone to technical anomalies that can undermine the trustworthiness of the data and the accuracy of statistical inferences, particularly in spatial and temporal analyses. Here we propose a framework for detecting anomalies in sensor data recorded in stream networks, which takes advantage of spatial and temporal autocorrelation to improve detection rates. The proposed framework involves the implementation of effective data imputation to handle missing data, alignment of time-series to address temporal disparities, and the identification of water quality events. We explore the effectiveness of a suite of state-of-the-art statistical methods including posterior predictive distributions, finite mixtures, and Hidden Markov Models (HMM). We showcase the practical implementation of automated anomaly detection in near-real time by employing a Bayesian recursive approach. This demonstration is conducted through a comprehensive simulation study and a practical application to a substantive case study situated in the Herbert River, located in Queensland, Australia, which flows into the Great Barrier Reef. We found that methods such as posterior predictive distributions and HMM produce the best performance in detecting multiple types of anomalies. Utilizing data from multiple sensors deployed relatively near one another enhances the ability to distinguish between water quality events and technical anomalies, thereby significantly improving the accuracy of anomaly detection. Thus, uncertainty and biases in water quality reporting, interpretation, and modelling are reduced, and the effectiveness of subsequent management actions improved.

The integration of renewable energy sources (RES) into power grids presents significant challenges due to their intrinsic stochasticity and uncertainty, necessitating the development of new techniques for reliable and efficient forecasting. This paper proposes a method combining probabilistic forecasting and Gaussian copula for day-ahead prediction and scenario generation of load, wind, and solar power in high-dimensional contexts. By incorporating weather covariates and restoring spatio-temporal correlations, the proposed method enhances the reliability of probabilistic forecasts in RES. Extensive numerical experiments compare the effectiveness of different time series models, with performance evaluated using comprehensive metrics on a real-world and high-dimensional dataset from Midcontinent Independent System Operator (MISO). The results highlight the importance of weather information and demonstrate the efficacy of the Gaussian copula in generating realistic scenarios, with the proposed weather-informed Temporal Fusion Transformer (WI-TFT) model showing superior performance.

Solar flare prediction studies have been recently conducted with the use of Space-Weather MDI (Michelson Doppler Imager onboard Solar and Heliospheric Observatory) Active Region Patches (SMARP) and Space-Weather HMI (Helioseismic and Magnetic Imager onboard Solar Dynamics Observatory) Active Region Patches (SHARP), which are two currently available data products containing magnetic field characteristics of solar active regions. The present work is an effort to combine them into one data product, and perform some initial statistical analyses in order to further expand their application in space weather forecasting. The combined data are derived by filtering, rescaling, and merging the SMARP with SHARP parameters, which can then be spatially reduced to create uniform multivariate time series. The resulting combined MDI-HMI dataset currently spans the period between April 4, 1996, and December 13, 2022, and may be extended to a more recent date. This provides an opportunity to correlate and compare it with other space weather time series, such as the daily solar flare index or the statistical properties of the soft X-ray flux measured by the Geostationary Operational Environmental Satellites (GOES). Time-lagged cross-correlation indicates that a relationship may exist, where some magnetic field properties of active regions lead the flare index in time. Applying the rolling window technique makes it possible to see how this leader-follower dynamic varies with time. Preliminary results indicate that areas of high correlation generally correspond to increased flare activity during the peak solar cycle.

Vector autoregressions (VARs) are popular model for analyzing multivariate economic time series. However, VARs can be over-parameterized if the numbers of variables and lags are moderately large. Tensor VAR, a recent solution to over-parameterization, treats the coefficient matrix as a third-order tensor and estimates the corresponding tensor decomposition to achieve parsimony. In this paper, we employ the Tensor VAR structure with a CANDECOMP/PARAFAC (CP) decomposition and conduct Bayesian inference to estimate parameters. Firstly, we determine the rank by imposing the Multiplicative Gamma Prior to the tensor margins, i.e. elements in the decomposition, and accelerate the computation with an adaptive inferential scheme. Secondly, to obtain interpretable margins, we propose an interweaving algorithm to improve the mixing of margins and identify the margins using a post-processing procedure. In an application to the US macroeconomic data, our models outperform standard VARs in point and density forecasting and yield a summary of the dynamic of the US economy.

Precise and timely forecasting of blood glucose levels is essential for effective diabetes management. While extensive research has been conducted on Type 1 diabetes mellitus, Type 2 diabetes mellitus (T2DM) presents unique challenges due to its heterogeneity, underscoring the need for specialized blood glucose forecasting systems. This study introduces a novel blood glucose forecasting system, applied to a dataset of 100 patients from the ShanghaiT2DM study. Our study uniquely integrates knowledge-driven and data-driven approaches, leveraging expert knowledge to validate and interpret the relationships among diabetes-related variables and deploying the data-driven approach to provide accurate forecast blood glucose levels. The Bayesian network approach facilitates the analysis of dependencies among various diabetes-related variables, thus enabling the inference of continuous glucose monitoring (CGM) trajectories in similar individuals with T2DM. By incorporating past CGM data including inference CGM trajectories, dietary records, and individual-specific information, the Bayesian structural time series (BSTS) model effectively forecasts glucose levels across time intervals ranging from 15 to 60 minutes. Forecast results show a mean absolute error of 6.41 mg/dL, a root mean square error of 8.29 mg/dL, and a mean absolute percentage error of 5.28%, for a 15-minute prediction horizon. This study makes the first application of the ShanghaiT2DM dataset for glucose level forecasting, considering the influences of diabetes-related variables. Its findings establish a foundational framework for developing personalized diabetes management strategies, potentially enhancing diabetes care through more accurate and timely interventions.

Estimation in GARMA models has traditionally been carried out under the frequentist approach. To date, Bayesian approaches for such estimation have been relatively limited. In the context of GARMA models for count time series, Bayesian estimation achieves satisfactory results in terms of point estimation. Model selection in this context often relies on the use of information criteria. Despite its prominence in the literature, the use of information criteria for model selection in GARMA models for count time series have been shown to present poor performance in simulations, especially in terms of their ability to correctly identify models, even under large sample sizes. In this study, we study the problem of order selection in GARMA models for count time series, adopting a Bayesian perspective through the application of the Reversible Jump Markov Chain Monte Carlo approach. Monte Carlo simulation studies are conducted to assess the finite sample performance of the developed ideas, including point and interval inference, sensitivity analysis, effects of burn-in and thinning, as well as the choice of related priors and hyperparameters. Two real-data applications are presented, one considering automobile production in Brazil and the other considering bus exportation in Brazil before and after the COVID-19 pandemic, showcasing the method's capabilities and further exploring its flexibility.

Modern vehicles are equipped with Electronic Control Units (ECU) that are used for controlling important vehicle functions including safety-critical operations. ECUs exchange information via in-vehicle communication buses, of which the Controller Area Network (CAN bus) is by far the most widespread representative. Problems that may occur in the vehicle's physical parts or malicious attacks may cause anomalies in the CAN traffic, impairing the correct vehicle operation. Therefore, the detection of such anomalies is vital for vehicle safety. This paper reviews the research on anomaly detection for in-vehicle networks, more specifically for the CAN bus. Our main focus is the evaluation of methods used for CAN bus anomaly detection together with the datasets used in such analysis. To provide the reader with a more comprehensive understanding of the subject, we first give a brief review of related studies on time series-based anomaly detection. Then, we conduct an extensive survey of recent deep learning-based techniques as well as conventional techniques for CAN bus anomaly detection. Our comprehensive analysis delves into anomaly detection algorithms employed in in-vehicle networks, specifically focusing on their learning paradigms, inherent strengths, and weaknesses, as well as their efficacy when applied to CAN bus datasets. Lastly, we highlight challenges and open research problems in CAN bus anomaly detection.

We propose a test to detect a forecast accuracy breakdown in a long memory time series and provide theoretical and simulation evidence on the memory transfer from the time series to the forecast residuals. The proposed method uses a double sup-Wald test against the alternative of a structural break in the mean of an out-of-sample loss series. To address the problem of estimating the long-run variance under long memory, a robust estimator is applied. The corresponding breakpoint results from a long memory robust CUSUM test. The finite sample size and power properties of the test are derived in a Monte Carlo simulation. A monotonic power function is obtained for the fixed forecasting scheme. In our practical application, we find that the global energy crisis that began in 2021 led to a forecast break in European electricity prices, while the results for the U.S. are mixed.

Kolmogorov-Arnold Networks (KAN) has recently attracted significant attention as a promising alternative to traditional Multi-Layer Perceptrons (MLP). Despite their theoretical appeal, KAN require validation on large-scale benchmark datasets. Time series data, which has become increasingly prevalent in recent years, especially univariate time series are naturally suited for validating KAN. Therefore, we conducted a fair comparison among KAN, MLP, and mixed structures. The results indicate that KAN can achieve performance comparable to, or even slightly better than, MLP across 128 time series datasets. We also performed an ablation study on KAN, revealing that the output is primarily determined by the base component instead of b-spline function. Furthermore, we assessed the robustness of these models and found that KAN and the hybrid structure MLP_KAN exhibit significant robustness advantages, attributed to their lower Lipschitz constants. This suggests that KAN and KAN layers hold strong potential to be robust models or to improve the adversarial robustness of other models.

Multimodal sentiment analysis aims to learn representations from different modalities to identify human emotions. However, existing works often neglect the frame-level redundancy inherent in continuous time series, resulting in incomplete modality representations with noise. To address this issue, we propose temporal-invariant learning for the first time, which constrains the distributional variations over time steps to effectively capture long-term temporal dynamics, thus enhancing the quality of the representations and the robustness of the model. To fully exploit the rich semantic information in textual knowledge, we propose a semantic-guided fusion module. By evaluating the correlations between different modalities, this module facilitates cross-modal interactions gated by modality-invariant representations. Furthermore, we introduce a modality discriminator to disentangle modality-invariant and modality-specific subspaces. Experimental results on two public datasets demonstrate the superiority of our model. Our code is available at https://github.com/X-G-Y/SATI.

change Title, Authors, Abstract

While it is widely recognised that linear (structural) VARs may fail to capture important aspects of economic time series, the use of nonlinear SVARs has to date been almost entirely confined to the modelling of stationary time series, because of a lack of understanding as to how common stochastic trends may be accommodated within nonlinear models. This has unfortunately circumscribed the range of series to which such models can be applied -- and/or required that these series be first transformed to stationarity, a potential source of misspecification -- and prevented the use of long-run identifying restrictions in these models. To address these problems, we develop a flexible class of additively time-separable nonlinear SVARs, which subsume models with threshold-type endogenous regime switching, both of the piecewise linear and smooth transition varieties. We extend the Granger--Johansen representation theorem to this class of models, obtaining conditions that specialise exactly to the usual ones when the model is linear. We further show that, as a corollary, these models are capable of supporting the same kinds of long-run identifying restrictions as are available in linearly cointegrated SVARs.

ii + 50 pp., 1 figure

Machine learning models play a vital role in time series forecasting. These models, however, often overlook an important element: point uncertainty estimates. Incorporating these estimates is crucial for effective risk management, informed model selection, and decision-making.To address this issue, our research introduces a method for uncertainty estimation. We employ a surrogate Gaussian process regression model. It enhances any base regression model with reasonable uncertainty estimates. This approach stands out for its computational efficiency. It only necessitates training one supplementary surrogate and avoids any data-specific assumptions. Furthermore, this method for work requires only the presence of the base model as a black box and its respective training data. The effectiveness of our approach is supported by experimental results. Using various time-series forecasting data, we found that our surrogate model-based technique delivers significantly more accurate confidence intervals. These techniques outperform both bootstrap-based and built-in methods in a medium-data regime. This superiority holds across a range of base model types, including a linear regression, ARIMA, gradient boosting and a neural network.

Time series forecasting models typically rely on a fixed-size training set and treat all data uniformly, which may not effectively capture the specific patterns present in more challenging training samples. To address this issue, we introduce AutoTSAug, a learnable data augmentation method based on reinforcement learning. Our approach begins with an empirical analysis to determine which parts of the training data should be augmented. Specifically, we identify the so-called marginal samples by considering the prediction diversity across a set of pretrained forecasting models. Next, we propose using variational masked autoencoders as the augmentation model and applying the REINFORCE algorithm to transform the marginal samples into new data. The goal of this generative model is not only to mimic the distribution of real data but also to reduce the variance of prediction errors across the model zoo. By augmenting the marginal samples with a learnable policy, AutoTSAug substantially improves forecasting performance, advancing the prior art in this field with minimal additional computational cost.

Multivariate time series forecasting relies on accurately capturing the correlations among variates. Current channel-independent (CI) models and models with a CI final projection layer are unable to capture these dependencies. In this paper, we present the variate embedding (VE) pipeline, which learns a unique and consistent embedding for each variate and combines it with Mixture of Experts (MoE) and Low-Rank Adaptation (LoRA) techniques to enhance forecasting performance while controlling parameter size. The VE pipeline can be integrated into any model with a CI final projection layer to improve multivariate forecasting. The learned VE effectively groups variates with similar temporal patterns and separates those with low correlations. The effectiveness of the VE pipeline is demonstrated through extensive experiments on four widely-used datasets. The code is available at: \url{https://github.com/swang-song/VE}.

We introduce a new framework to analyze shape descriptors that capture the geometric features of an ensemble of point clouds. At the core of our approach is the point of view that the data arises as sampled recordings from a metric space-valued stochastic process, possibly of nonstationary nature, thereby integrating geometric data analysis into the realm of functional time series analysis. Our framework allows for natural incorporation of spatial-temporal dynamics, heterogeneous sampling, and the study of convergence rates. Further, we derive complete invariants for classes of metric space-valued stochastic processes in the spirit of Gromov, and relate these invariants to so-called ball volume processes. Under mild dependence conditions, a weak invariance principle in $D([0,1]\times [0,\mathscr{R}])$ is established for sequential empirical versions of the latter, assuming the probabilistic structure possibly changes over time. Finally, we use this result to introduce novel test statistics for topological change, which are distribution-free in the limit under the hypothesis of stationarity. We explore these test statistics on time series of single-cell mRNA expression data, using shape descriptors coming from topological data analysis.

The present paper is concerned with deep learning techniques applied to detection and localization of damage in a thin aluminum plate. We used data generated on a tabletop apparatus by mounting to the plate four piezoelectric transducers, each of which took turn to generate a Lamb wave that then traversed the region of interest before being received by the remaining three sensors. On training a neural network to analyze time-series data of the material response, which displayed damage-reflective features whenever the plate guided waves interacted with a contact load, we achieved a model that detected with greater than 99% accuracy in addition to a model that localized with $3.14 \pm 0.21$ mm mean distance error and captured more than 60% of test examples within the diffraction limit. For each task, the best-performing model was designed according to the inductive bias that our transducers were both similar and arranged in a square pattern on a nearly uniform plate.

Accurate prediction of pedestrian trajectories is crucial for improving the safety of autonomous driving. However, this task is generally nontrivial due to the inherent stochasticity of human motion, which naturally requires the predictor to generate multi-modal prediction. Previous works leverage various generative methods, such as GAN and VAE, for pedestrian trajectory prediction. Nevertheless, these methods may suffer from mode collapse and relatively low-quality results. The denoising diffusion probabilistic model (DDPM) has recently been applied to trajectory prediction due to its simple training process and powerful reconstruction ability. However, current diffusion-based methods do not fully utilize input information and usually require many denoising iterations that lead to a long inference time or an additional network for initialization. To address these challenges and facilitate the use of diffusion models in multi-modal trajectory prediction, we propose GDTS, a novel Goal-Guided Diffusion Model with Tree Sampling for multi-modal trajectory prediction. Considering the "goal-driven" characteristics of human motion, GDTS leverages goal estimation to guide the generation of the diffusion network. A two-stage tree sampling algorithm is presented, which leverages common features to reduce the inference time and improve accuracy for multi-modal prediction. Experimental results demonstrate that our proposed framework achieves comparable state-of-the-art performance with real-time inference speed in public datasets.

Submitted to ICRA 2025

This paper presents a generic trajectory planning method for wheeled robots with fixed steering axes while the steering angle of each wheel is constrained. In the existing literatures, All-Wheel-Steering (AWS) robots, incorporating modes such as rotation-free translation maneuvers, in-situ rotational maneuvers, and proportional steering, exhibit inefficient performance due to time-consuming mode switches. This inefficiency arises from wheel rotation constraints and inter-wheel cooperation requirements. The direct application of a holonomic moving strategy can lead to significant slip angles or even structural failure. Additionally, the limited steering range of AWS wheeled robots exacerbates non-linearity characteristics, thereby complicating control processes. To address these challenges, we developed a novel planning method termed Constrained AWS (C-AWS), which integrates second-order discrete search with predictive control techniques. Experimental results demonstrate that our method adeptly generates feasible and smooth trajectories for C-AWS while adhering to steering angle constraints.

Accepted by iROS 2024

Predicting crowded intents and trajectories is crucial in varouls real-world applications, including service robots and autonomous vehicles. Understanding environmental dynamics is challenging, not only due to the complexities of modeling pair-wise spatial and temporal interactions but also the diverse influence of group-wise interactions. To decode the comprehensive pair-wise and group-wise interactions in crowded scenarios, we introduce Hyper-STTN, a Hypergraph-based Spatial-Temporal Transformer Network for crowd trajectory prediction. In Hyper-STTN, crowded group-wise correlations are constructed using a set of multi-scale hypergraphs with varying group sizes, captured through random-walk robability-based hypergraph spectral convolution. Additionally, a spatial-temporal transformer is adapted to capture pedestrians' pair-wise latent interactions in spatial-temporal dimensions. These heterogeneous group-wise and pair-wise are then fused and aligned though a multimodal transformer network. Hyper-STTN outperformes other state-of-the-art baselines and ablation models on 5 real-world pedestrian motion datasets.

Social media platforms enable users to share diverse types of information, including geolocation data that captures their movement patterns. Such geolocation data can be leveraged to reconstruct the trajectory of a user's visited Points of Interest (POIs). A key requirement in numerous applications is the ability to measure the similarity between such trajectories, as this facilitates the retrieval of trajectories that are similar to a given reference trajectory. This is the main focus of our work. Existing methods predominantly rely on applying a similarity function to each candidate trajectory to identify those that are sufficiently similar. However, this approach becomes computationally expensive when dealing with large-scale datasets. To mitigate this challenge, we propose TISIS, an efficient method that uses trajectory indexing to quickly find similar trajectories that share common POIs in the same order. Furthermore, to account for scenarios where POIs in trajectories may not exactly match but are contextually similar, we introduce TISIS*, a variant of TISIS that incorporates POI embeddings. This extension allows for more comprehensive retrieval of similar trajectories by considering semantic similarities between POIs, beyond mere exact matches. Extensive experimental evaluations demonstrate that the proposed approach significantly outperforms a baseline method based on the well-known Longest Common SubSequence (LCSS) algorithm, yielding substantial performance improvements across various real-world datasets.

Tracking controllers enable robotic systems to accurately follow planned reference trajectories. In particular, reinforcement learning (RL) has shown promise in the synthesis of controllers for systems with complex dynamics and modest online compute budgets. However, the poor sample efficiency of RL and the challenges of reward design make training slow and sometimes unstable, especially for high-dimensional systems. In this work, we leverage the inherent Lie group symmetries of robotic systems with a floating base to mitigate these challenges when learning tracking controllers. We model a general tracking problem as a Markov decision process (MDP) that captures the evolution of both the physical and reference states. Next, we prove that symmetry in the underlying dynamics and running costs leads to an MDP homomorphism, a mapping that allows a policy trained on a lower-dimensional "quotient" MDP to be lifted to an optimal tracking controller for the original system. We compare this symmetry-informed approach to an unstructured baseline, using Proximal Policy Optimization (PPO) to learn tracking controllers for three systems: the Particle (a forced point mass), the Astrobee (a fullyactuated space robot), and the Quadrotor (an underactuated system). Results show that a symmetry-aware approach both accelerates training and reduces tracking error after the same number of training steps.

The first three authors contributed equally to this work

Semi-supervised 3D object detection is a common strategy employed to circumvent the challenge of manually labeling large-scale autonomous driving perception datasets. Pseudo-labeling approaches to semi-supervised learning adopt a teacher-student framework in which machine-generated pseudo-labels on a large unlabeled dataset are used in combination with a small manually-labeled dataset for training. In this work, we address the problem of improving pseudo-label quality through leveraging long-term temporal information captured in driving scenes. More specifically, we leverage pre-trained motion-forecasting models to generate object trajectories on pseudo-labeled data to further enhance the student model training. Our approach improves pseudo-label quality in two distinct manners: first, we suppress false positive pseudo-labels through establishing consistency across multiple frames of motion forecasting outputs. Second, we compensate for false negative detections by directly inserting predicted object tracks into the pseudo-labeled scene. Experiments on the nuScenes dataset demonstrate the effectiveness of our approach, improving the performance of standard semi-supervised approaches in a variety of settings.

Future spacecraft and surface robotic missions require increasingly capable autonomy stacks for exploring challenging and unstructured domains, and trajectory optimization will be a cornerstone of such autonomy stacks. However, the nonlinear optimization solvers required remain too slow for use on relatively resource-constrained flight-grade computers. In this work, we turn towards amortized optimization, a learning-based technique for accelerating optimization run times, and present TOAST: Trajectory Optimization with Merit Function Warm Starts. Offline, using data collected from a simulation, we train a neural network to learn a mapping to the full primal and dual solutions given the problem parameters. Crucially, we build upon recent results from decision-focused learning and present a set of decision-focused loss functions using the notion of merit functions for optimization problems. We show that training networks with such constraint-informed losses can better encode the structure of the trajectory optimization problem and jointly learn to reconstruct the primal-dual solution while yielding improved constraint satisfaction. Through numerical experiments on a Lunar rover problem and a 3-degrees-of-freedom Mars powered descent guidance problem, we demonstrate that TOAST outperforms benchmark approaches in terms of both computation times and network prediction constraint satisfaction.

Advances in medical imaging technologies have enabled the collection of longitudinal images, which involve repeated scanning of the same patients over time, to monitor disease progression. However, predictive modeling of such data remains challenging due to high dimensionality, irregular sampling, and data sparsity. To address these issues, we propose ImageFlowNet, a novel model designed to forecast disease trajectories from initial images while preserving spatial details. ImageFlowNet first learns multiscale joint representation spaces across patients and time points, then optimizes deterministic or stochastic flow fields within these spaces using a position-parameterized neural ODE/SDE framework. The model leverages a UNet architecture to create robust multiscale representations and mitigates data scarcity by combining knowledge from all patients. We provide theoretical insights that support our formulation of ODEs, and motivate our regularizations involving high-level visual features, latent space organization, and trajectory smoothness. We validate ImageFlowNet on three longitudinal medical image datasets depicting progression in geographic atrophy, multiple sclerosis, and glioblastoma, demonstrating its ability to effectively forecast disease progression and outperform existing methods. Our contributions include the development of ImageFlowNet, its theoretical underpinnings, and empirical validation on real-world datasets. The official implementation is available at https://github.com/KrishnaswamyLab/ImageFlowNet.

Updated narration and moved ablation to main text

In this work, we introduce a novel gradient descent-based approach for optimizing control systems, leveraging a new representation of stable closed-loop dynamics as a function of two matrices i.e. the step size or direction matrix and value matrix of the Lyapunov cost function. This formulation provides a new framework for analyzing and designing feedback control laws. We show that any stable closed-loop system can be expressed in this form with appropriate values for the step size and value matrices. Furthermore, we show that this parameterization of the closed-loop system is equivalent to a linear quadratic regulator for appropriately chosen weighting matrices. We also show that trajectories can be shaped using this approach to achieve a desired closed-loop behavior.

Marginal structural models (MSMs) are widely used in observational studies to estimate the causal effect of time-varying treatments. Despite its popularity, limited attention has been paid to summarizing the treatment history in the outcome model, which proves particularly challenging when individuals' treatment trajectories exhibit complex patterns over time. Commonly used metrics such as the average treatment level fail to adequately capture the treatment history, hindering causal interpretation. For scenarios where treatment histories exhibit distinct temporal patterns, we develop a new approach to parameterize the outcome model. We apply latent growth curve analysis to identify representative treatment trajectories from the observed data and use the posterior probability of latent class membership to summarize the different treatment trajectories. We demonstrate its use in parameterizing the MSMs, which facilitates the interpretations of the results. We apply the method to analyze data from an existing cohort of lung transplant recipients to estimate the effect of Tacrolimus concentrations on the risk of incident chronic kidney disease.

We have discovered that the core idea of our paper overlaps with a previously published work. In light of this, we need to conduct a more thorough update and revision of our research before proceeding further

Decision-making, motion planning, and trajectory prediction are crucial in autonomous driving systems. By accurately forecasting the movements of other road users, the decision-making capabilities of the autonomous system can be enhanced, making it more effective in responding to dynamic and unpredictable environments and more adaptive to diverse road scenarios. This paper presents the FFStreams++ approach for decision-making and motion planning of different maneuvers, including unprotected left turn, overtaking, and keep-lane. FFStreams++ is a combination of sampling-based and search-based approaches, where iteratively new sampled trajectories for different maneuvers are generated and optimized, and afterward, a heuristic search planner is called, searching for an optimal plan. We model the autonomous diving system in the Planning Domain Definition Language (PDDL) and search for the optimal plan using a heuristic Fast-Forward planner. In this approach, the initial state of the problem is modified iteratively through streams, which will generate maneuver-specific trajectory candidates, increasing the iterating level until an optimal plan is found. FFStreams++ integrates a query-connected network model for predicting possible future trajectories for each surrounding obstacle along with their probabilities. The proposed approach was tested on the CommonRoad simulation framework. We use a collection of randomly generated driving scenarios for overtaking and unprotected left turns at intersections to evaluate the FFStreams++ planner. The test results confirmed that the proposed approach can effectively execute various maneuvers to ensure safety and reduce the risk of collisions with nearby traffic agents.

This work is concerned with developing a data-driven approach for learning control barrier certificates (CBCs) and associated safety controllers for discrete-time nonlinear polynomial systems with unknown mathematical models, guaranteeing system safety over an infinite time horizon. The proposed approach leverages measured data acquired through an input-output observation, referred to as a single trajectory, collected over a specified time horizon. By fulfilling a certain rank condition, which ensures the unknown system is persistently excited by the collected data, we design a CBC and its corresponding safety controller directly from the finite-length observed data, without explicitly identifying the unknown dynamical system. This is achieved through proposing a data-based sum-of-squares optimization (SOS) program to systematically design CBCs and their safety controllers. We validate our data-driven approach over two physical case studies including a jet engine and a Lorenz system, demonstrating the efficacy of our proposed method.

We present Model Predictive Planning (MPP), a trajectory planner for low-agility vehicles such as a fixed-wing aircraft to navigate obstacle-laden environments. MPP consists of (1) a multi-path planning procedure that identifies candidate paths, (2) a raytracing procedure that generates linear constraints around these paths to enforce obstacle avoidance, and (3) a convex quadratic program that finds a feasible trajectory within these constraints if one exists. Low-agility aircraft cannot track arbitrary paths, so refining a given path into a trajectory that respects the vehicle's limited maneuverability and avoids obstacles often leads to an infeasible optimization problem. The critical feature of MPP is that it efficiently considers multiple candidate paths during the refinement process, thereby greatly increasing the chance of finding a feasible and trackable trajectory. We demonstrate the effectiveness of MPP on a longitudinal aircraft model.

The idea of generalized indices is one of the success stories of database systems research. It has found its way to implementation in common database systems. GiST (Generalized Search Tree) and SP-GiST (Space-Partitioned Generalized Search Tree) are two widely-used generalized indices that are typically used for multidimensional data. Currently, the generalized indices GiST and SP-GiST represent one database object using one index entry, e.g., a bounding box for each spatio-temporal object. However, when dealing with complex objects, e.g., moving object trajectories, a single entry per object is inadequate for creating efficient indices. Previous research has highlighted that splitting trajectories into multiple bounding boxes prior to indexing can enhance query performance as it leads to a higher index filter. In this paper, we introduce MGiST and MSP-GiST, the multi-entry generalized search tree counterparts of GiST and SP-GiST, respectively, that are designed to enable the partitioning of objects into multiple entries during insertion. The methods for decomposing a complex object into multiple sub-objects differ from one data type to another, and may depend on some domain-specific parameters. Thus, MGiST and MSP-GiST are designed to allow for pluggable modules that aid in optimizing the split of an object into multiple sub-objects. We demonstrate the usefulness of MGiST and MSP-GiST using a trajectory indexing scenario, where we realize several trajectory indexes using MGiST and MSP-GiST and instantiate these search trees with trajectory-specific splitting algorithms. We create and test the performance of several multi-entry versions of widely-used spatial index structures, e.g., R-Tree, Quad-Tree, and KD-Tree. We conduct evaluations using both synthetic and real-world data, and observe up to an order of magnitude enhancement in performance of point, range, and KNN queries.

Reusing pre-collected data from different domains is an attractive solution in decision-making tasks where the accessible data is insufficient in the target domain but relatively abundant in other related domains. Existing cross-domain policy transfer methods mostly aim at learning domain correspondences or corrections to facilitate policy learning, which requires learning domain/task-specific model components, representations, or policies that are inflexible or not fully reusable to accommodate arbitrary domains and tasks. These issues make us wonder: can we directly bridge the domain gap at the data (trajectory) level, instead of devising complicated, domain-specific policy transfer models? In this study, we propose a Cross-Domain Trajectory EDiting (xTED) framework with a new diffusion transformer model (Decision Diffusion Transformer, DDiT) that captures the trajectory distribution from the target dataset as a prior. The proposed diffusion transformer backbone captures the intricate dependencies among state, action, and reward sequences, as well as the transition dynamics within the target data trajectories. With the above pre-trained diffusion prior, source data trajectories with domain gaps can be transformed into edited trajectories that closely resemble the target data distribution through the diffusion-based editing process, which implicitly corrects the underlying domain gaps, enhancing the state realism and dynamics reliability in source trajectory data, while enabling flexible choices of downstream policy learning methods. Despite its simplicity, xTED demonstrates superior performance against other baselines in extensive simulation and real-robot experiments.

xTED offers a novel, generic, flexible, simple and effective paradigm that casts cross-domain policy adaptation as a data pre-processing problem

This paper presents SPI-DP, a novel first-order optimizer capable of optimizing robot programs with respect to both high-level task objectives and motion-level constraints. To that end, we introduce DGPMP2-ND, a differentiable collision-free motion planner for serial N-DoF kinematics, and integrate it into an iterative, gradient-based optimization approach for generic, parameterized robot program representations. SPI-DP allows first-order optimization of planned trajectories and program parameters with respect to objectives such as cycle time or smoothness subject to e.g. collision constraints, while enabling humans to understand, modify or even certify the optimized programs. We provide a comprehensive evaluation on two practical household and industrial applications.

8 pages, 6 figures, submitted to the 2025 IEEE International Conference on Robotics & Automation (ICRA)

Non-prehensile manipulation in high-dimensional systems is challenging for a variety of reasons. One of the main reasons is the computationally long planning times that come with a large state space. Trajectory optimisation algorithms have proved their utility in a wide variety of tasks, but, like most methods struggle scaling to the high dimensional systems ubiquitous to non-prehensile manipulation in clutter as well as deformable object manipulation. We reason that, during manipulation, different degrees of freedom will become more or less important to the task over time as the system evolves. We leverage this idea to reduce the number of degrees of freedom considered in a trajectory optimisation problem, to reduce planning times. This idea is particularly relevant in the context of model predictive control (MPC) where the cost landscape of the optimisation problem is constantly evolving. We provide simulation results under asynchronous MPC and show our methods are capable of achieving better overall performance due to the decreased policy lag whilst still being able to optimise trajectories effectively.

18 pages, 4 figures, accepted to WAFR 2024

Differential-driven robots are widely used in various scenarios thanks to their straightforward principle, from household service robots to disaster response field robots. There are several different types of deriving mechanisms considering the real-world applications, including two-wheeled, four-wheeled skid-steering, tracked robots, etc. The differences in the driving mechanism usually require specific kinematic modeling when precise controlling is desired. Furthermore, the nonholonomic dynamics and possible lateral slip lead to different degrees of difficulty in getting feasible and high-quality trajectories. Therefore, a comprehensive trajectory optimization framework to compute trajectories efficiently for various kinds of differential-driven robots is highly desirable. In this paper, we propose a universal trajectory optimization framework that can be applied to differential-driven robot class, enabling the generation of high-quality trajectories within a restricted computational timeframe. We introduce a novel trajectory representation based on polynomial parameterization of motion states or their integrals, such as angular and linear velocities, that inherently matching robots' motion to the control principle for differential-driven robot class. The trajectory optimization problem is formulated to minimize complexity while prioritizing safety and operational efficiency. We then build a full-stack autonomous planning and control system to show the feasibility and robustness. We conduct extensive simulations and real-world testing in crowded environments with three kinds of differential-driven robots to validate the effectiveness of our approach. We will release our method as an open-source package.

Deep neural networks (DNN) are increasingly being used to learn controllers due to their excellent approximation capabilities. However, their black-box nature poses significant challenges to closed-loop stability guarantees and performance analysis. In this paper, we introduce a structured DNN-based controller for the trajectory tracking control of Lagrangian systems using backing techniques. By properly designing neural network structures, the proposed controller can ensure closed-loop stability for any compatible neural network parameters. In addition, improved control performance can be achieved by further optimizing neural network parameters. Besides, we provide explicit upper bounds on tracking errors in terms of controller parameters, which allows us to achieve the desired tracking performance by properly selecting the controller parameters. Furthermore, when system models are unknown, we propose an improved Lagrangian neural network (LNN) structure to learn the system dynamics and design the controller. We show that in the presence of model approximation errors and external disturbances, the closed-loop stability and tracking control performance can still be guaranteed. The effectiveness of the proposed approach is demonstrated through simulations.

We present AdaFold, a model-based feedback-loop framework for optimizing folding trajectories. AdaFold extracts a particle-based representation of cloth from RGB-D images and feeds back the representation to a model predictive control to replan folding trajectory at every time step. A key component of AdaFold that enables feedback-loop manipulation is the use of semantic descriptors extracted from geometric features. These descriptors enhance the particle representation of the cloth to distinguish between ambiguous point clouds of differently folded cloths. Our experiments demonstrate AdaFold's ability to adapt folding trajectories of cloths with varying physical properties and generalize from simulated training to real-world execution.

8 pages, 6 figures, 5 tables

Learning rewards from human behaviour or feedback is a promising approach to aligning AI systems with human values but fails to consistently extract correct reward functions. Interpretability tools could enable users to understand and evaluate possible flaws in learned reward functions. We propose Counterfactual Trajectory Explanations (CTEs) to interpret reward functions in reinforcement learning by contrasting an original with a counterfactual partial trajectory and the rewards they each receive. We derive six quality criteria for CTEs and propose a novel Monte-Carlo-based algorithm for generating CTEs that optimises these quality criteria. Finally, we measure how informative the generated explanations are to a proxy-human model by training it on CTEs. CTEs are demonstrably informative for the proxy-human model, increasing the similarity between its predictions and the reward function on unseen trajectories. Further, it learns to accurately judge differences in rewards between trajectories and generalises to out-of-distribution examples. Although CTEs do not lead to a perfect understanding of the reward, our method, and more generally the adaptation of XAI methods, are presented as a fruitful approach for interpreting learned reward functions.

Trajectory data is essential for various applications as it records the movement of vehicles. However, publicly available trajectory datasets remain limited in scale due to privacy concerns, which hinders the development of trajectory data mining and trajectory-based applications. To address this issue, some methods for generating synthetic trajectories have been proposed to expand the scale of the dataset. However, all existing methods generate trajectories in the geographical coordinate system, which poses two limitations for their utilization in practical applications: 1) the inability to ensure that the generated trajectories are constrained on the road. 2) the lack of road-related information. In this paper, we propose a new problem to meet the practical application need, \emph{i.e.}, road network-constrained trajectory (RNTraj) generation, which can directly generate trajectories on the road network with road-related information. RNTraj is a hybrid type of data, in which each point is represented by a discrete road segment and a continuous moving rate. To generate RNTraj, we design a diffusion model called Diff-RNTraj. This model can effectively handle the hybrid RNTraj using a continuous diffusion framework by incorporating a pre-training strategy to embed hybrid RNTraj into continuous representations. During the sampling stage, a RNTraj decoder is designed to map the continuous representation generated by the diffusion model back to the hybrid RNTraj format. Furthermore, Diff-RNTraj introduces a novel loss function to enhance the spatial validity of the generated trajectories. Extensive experiments conducted on two real-world trajectory datasets demonstrate the effectiveness of the proposed model.

This paper has been accepted as a regular paper at IEEE TKDE

Network inference, the task of reconstructing interactions in a complex system from experimental observables, is a central yet extremely challenging problem in systems biology. While much progress has been made in the last two decades, network inference remains an open problem. For systems observed at steady state, limited insights are available since temporal information is unavailable and thus causal information is lost. Two common avenues for gaining causal insights into system behaviour are to leverage temporal dynamics in the form of trajectories, and to apply interventions such as knock-out perturbations. We propose an approach for leveraging both dynamical and perturbational single cell data to jointly learn cellular trajectories and power network inference. Our approach is motivated by min-entropy estimation for stochastic dynamics and can infer directed and signed networks from time-stamped single cell snapshots.

Simulating human mobility data is essential for various application domains, including transportation, urban planning, and epidemic control, since real data are often inaccessible to researchers due to expensive costs and privacy issues. Several existing deep generative solutions propose learning from real trajectories to generate synthetic ones. Despite the progress, most of them suffer from training stability issues and scale poorly with growing data size. More importantly, they generally lack control mechanisms to steer the generated trajectories based on spatiotemporal constraints such as fixing specific visits. To address such limitations, we formally define the controlled trajectory generation problem with spatiotemporal constraints and propose Geo-Llama. This novel LLM-inspired framework enforces explicit visit constraints in a contextually coherent way. It fine-tunes pre-trained LLMs on trajectories with a visit-wise permutation strategy where each visit corresponds to a time and location. This enables the model to capture the spatiotemporal patterns regardless of visit orders and allows flexible and in-context constraint integration through prompts during generation. Extensive experiments on real-world and synthetic datasets validate the effectiveness of Geo-Llama, demonstrating its versatility and robustness in handling a broad range of constraints to generate more realistic trajectories compared to existing methods.

One of the limitations of the commonly used Absolute Trajectory Error (ATE) is that it is highly sensitive to outliers. As a result, in the presence of just a few outliers, it often fails to reflect the varying accuracy as the inlier trajectory error or the number of outliers varies. In this work, we propose an alternative error metric for evaluating the accuracy of the reconstructed camera trajectory. Our metric, named Discernible Trajectory Error (DTE), is computed in five steps: (1) Shift the ground-truth and estimated trajectories such that both of their geometric medians are located at the origin. (2) Rotate the estimated trajectory such that it minimizes the sum of geodesic distances between the corresponding camera orientations. (3) Scale the estimated trajectory such that the median distance of the cameras to their geometric median is the same as that of the ground truth. (4) Compute, winsorize and normalize the distances between the corresponding cameras. (5) Obtain the DTE by taking the average of the mean and the root-mean-square (RMS) of the resulting distances. This metric is an attractive alternative to the ATE, in that it is capable of discerning the varying trajectory accuracy as the inlier trajectory error or the number of outliers varies. Using the similar idea, we also propose a novel rotation error metric, named Discernible Rotation Error (DRE), which has similar advantages to the DTE. Furthermore, we propose a simple yet effective method for calibrating the camera-to-marker rotation, which is needed for the computation of our metrics. Our methods are verified through extensive simulations.

The main part of this manuscript (except the part on DRE) has been accepted to ECCV 2024 Workshop HALF-CENTURY OF STRUCTURE-FROM-MOTION (50SFM)

This paper proposes an approach for controlling surgical robotic systems, while complying with the Remote Center of Motion (RCM) constraint in Robot-Assisted Minimally Invasive Surgery (RA-MIS). In this approach, the RCM-constraint is upheld algorithmically, providing flexibility in the positioning of the insertion point and enabling compatibility with a wide range of general-purpose robots. The paper further investigates the impact of the tool's insertion ratio on the RCM-error, and introduces a manipulability index of the robot which considers the RCM-error that it is used to find a starting configuration. To accurately evaluate the proposed method's trajectory tracking within an RCM-constrained environment, an electromagnetic tracking system is employed. The results demonstrate the effectiveness of the proposed method in addressing the RCM constraint problem in RA-MIS.

6 pages, 7 figures. Published in the 21st International Conference on Advanced Robotics (ICAR 2023)

Learning from Demonstration allows robots to mimic human actions. However, these methods do not model constraints crucial to ensure safety of the learned skill. Moreover, even when explicitly modelling constraints, they rely on the assumption of a known cost function, which limits their practical usability for task with unknown cost. In this work we propose a two-step optimization process that allow to estimate cost and constraints by decoupling the learning of cost functions from the identification of unknown constraints within the demonstrated trajectories. Initially, we identify the cost function by isolating the effect of constraints on parts of the demonstrations. Subsequently, a constraint leaning method is used to identify the unknown constraints. Our approach is validated both on simulated trajectories and a real robotic manipulation task. Our experiments show the impact that incorrect cost estimation has on the learned constraints and illustrate how the proposed method is able to infer unknown constraints, such as obstacles, from demonstrated trajectories without any initial knowledge of the cost.

(Accepted/In press) 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Many of the existing works on quadrotor control address the trajectory tracking problem by employing a cascade design in which the translational and rotational dynamics are stabilized by two separate controllers. The stability of the cascade is often proved by employing trajectory-based arguments, most notably, integral input-to-state stability. In this paper, we follow a different route and present a control law ensuring that a composite function constructed from the translational and rotational tracking errors is a Lyapunov function for the closed-loop cascade. In particular, starting from a generic control law for the double integrator, we develop a suitable attitude control extension, by leveraging a backstepping-like procedure. Using this construction, we provide an almost global stability certificate. The proposed design employs the unit sphere $\mathcal{S}^2$ to describe the rotational degrees of freedom required for position control. This enables a simpler controller tuning and an improved tracking performance with respect to previous global solutions. The new design is demonstrated via numerical simulations and on real-world experiments.

The problem of generalization in learning from demonstration (LfD) has received considerable attention over the years, particularly within the context of movement primitives, where a number of approaches have emerged. Recently, two important approaches have gained recognition. While one leverages via-points to adapt skills locally by modulating demonstrated trajectories, another relies on so-called task-parameterized models that encode movements with respect to different coordinate systems, using a product of probabilities for generalization. While the former are well-suited to precise, local modulations, the latter aim at generalizing over large regions of the workspace and often involve multiple objects. Addressing the quality of generalization by leveraging both approaches simultaneously has received little attention. In this work, we propose an interactive imitation learning framework that simultaneously leverages local and global modulations of trajectory distributions. Building on the kernelized movement primitives (KMP) framework, we introduce novel mechanisms for skill modulation from direct human corrective feedback. Our approach particularly exploits the concept of via-points to incrementally and interactively 1) improve the model accuracy locally, 2) add new objects to the task during execution and 3) extend the skill into regions where demonstrations were not provided. We evaluate our method on a bearing ring-loading task using a torque-controlled, 7-DoF, DLR SARA robot.

With the rapid development of drone technology, accurate detection of Unmanned Aerial Vehicles (UAVs) has become essential for applications such as surveillance, security, and airspace management. In this paper, we propose a novel trajectory-guided method, the Patch Intensity Convergence (PIC) technique, which generates high-fidelity bounding boxes for UAV detection tasks and no need for the effort required for labeling. The PIC technique forms the foundation for developing UAVDB, a database explicitly created for UAV detection. Unlike existing datasets, which often use low-resolution footage or focus on UAVs in simple backgrounds, UAVDB employs high-resolution video to capture UAVs at various scales, ranging from hundreds of pixels to nearly single-digit sizes. This broad-scale variation enables comprehensive evaluation of detection algorithms across different UAV sizes and distances. Applying the PIC technique, we can also efficiently generate detection datasets from trajectory or positional data, even without size information. We extensively benchmark UAVDB using YOLOv8 series detectors, offering a detailed performance analysis. Our findings highlight UAVDB's potential as a vital database for advancing UAV detection, particularly in high-resolution and long-distance tracking scenarios.

7 pages, 5 figures, 3 tables

End-to-end approaches with Reinforcement Learning (RL) and Imitation Learning (IL) have gained increasing popularity in autonomous driving. However, they do not involve explicit reasoning like classic robotics workflow, nor planning with horizons, leading strategies implicit and myopic. In this paper, we introduce our trajectory planning method that uses Behavioral Cloning (BC) for path-tracking and Proximal Policy Optimization (PPO) bootstrapped by BC for static obstacle nudging. It outputs lateral offset values to adjust the given reference trajectory, and performs modified path for different controllers. Our experimental results show that the algorithm can do path-tracking that mimics the expert performance, and avoiding collision to fixed obstacles by trial and errors. This method makes a good attempt at planning with learning-based methods in trajectory planning problems of autonomous driving.

In prioritized planning for vehicles, vehicles plan trajectories in parallel or in sequence. Parallel prioritized planning offers approximately consistent computation time regardless of the number of vehicles but struggles to guarantee collision-free trajectories. Conversely, sequential prioritized planning can guarantee collision-freeness but results in increased computation time as the number of sequentially computing vehicles, which we term computation levels, grows. This number is determined by the directed coupling graph resulted from the coupling and prioritization of vehicles. In this work, we guarantee safe trajectories in parallel planning through reachability analysis. Although these trajectories are collision-free, they tend to be conservative. We address this by planning with a subset of vehicles in sequence. We formulate the problem of selecting this subset as a graph partitioning problem that allows us to independently set computation levels. Our simulations demonstrate a reduction in computation levels by approximately 64% compared to sequential prioritized planning while maintaining the solution quality.

8 pages, 4 figures. This is an extended abstract of our previous work published at the 2024 European Control Conference (ECC), June 25-28, 2024. Stockholm, Sweden

Existing methods utilizing spatial information for sound source separation require prior knowledge of the direction of arrival (DOA) of the source or utilize estimated but imprecise localization results, which impairs the separation performance, especially when the sound sources are moving. In fact, sound source localization and separation are interconnected problems, that is, sound source localization facilitates sound separation while sound separation contributes to more precise source localization. This paper proposes a method utilizing the mutual facilitation mechanism between sound source localization and separation for moving sources. Initially, sound separation is conducted using rough preliminary sound source tracking results. Sound source tracking is then performed on the separated signals thus the tracking results can become more precise. The precise trajectory can further enhances the separation performance. This mutual facilitation process can be performed over several iterations. Simulation experiments conducted under reverberation conditions and with moving sound sources demonstrate that the proposed method can achieve more accurate separation based on more precise tracking results.

9 pages,7 figures,submitted to IEEE/ACM Transactions on Audio, Speech and Language Processing(TASLP)

Autonomous systems that efficiently utilize tools can assist humans in completing many common tasks such as cooking and cleaning. However, current systems fall short of matching human-level of intelligence in terms of adapting to novel tools. Prior works based on affordance often make strong assumptions about the environments and cannot scale to more complex, contact-rich tasks. In this work, we tackle this challenge and explore how agents can learn to use previously unseen tools to manipulate deformable objects. We propose to learn a generative model of the tool-use trajectories as a sequence of tool point clouds, which generalizes to different tool shapes. Given any novel tool, we first generate a tool-use trajectory and then optimize the sequence of tool poses to align with the generated trajectory. We train a single model on four different challenging deformable object manipulation tasks, using demonstration data from only one tool per task. The model generalizes to various novel tools, significantly outperforming baselines. We further test our trained policy in the real world with unseen tools, where it achieves the performance comparable to human. Additional materials can be found on our project website: https://sites.google.com/view/toolgen.

Currently, the spray-painting robot trajectory planning technology aiming at spray painting quality mainly applies to single-color spraying. Conventional methods of optimizing the spray gun trajectory based on simulated thickness can only qualitatively reflect the color distribution, and can not simulate the color effect of spray painting at the pixel level. Therefore, it is not possible to accurately control the area covered by the color and the gradation of the edges of the area, and it is also difficult to deal with the situation where multiple colors of paint are sprayed in combination. To solve the above problems, this paper is inspired by the Kubelka-Munk model and combines the 3D machine vision method and artificial neural network to propose a spray painting color effect prediction method. The method is enabled to predict the execution effect of the spray gun trajectory with pixel-level accuracy from the dimension of the surface color of the workpiece after spray painting. On this basis, the method can be used to replace the traditional thickness simulation method to establish the objective function of the spray gun trajectory optimization problem, and thus solve the difficult problem of spray gun trajectory optimization for multi-color paint combination spraying. In this paper, the mathematical model of the spray painting color effect prediction problem is first determined through the analysis of the Kubelka-Munk paint film color rendering model, and at the same time, the spray painting color effect dataset is established with the help of the depth camera and point cloud processing algorithm. After that, the multilayer perceptron model was improved with the help of gating and residual structure and was used for the color prediction task. To verify ...

High-precision surface defect detection in manufacturing is essential for ensuring quality control. Laser triangulation profilometric sensors are key to this process, providing detailed and accurate surface measurements over a line. To achieve a complete and precise surface scan, accurate relative motion between the sensor and the workpiece is required. It is crucial to control the sensor pose to maintain optimal distance and relative orientation to the surface. It is also important to ensure uniform profile distribution throughout the scanning process. This paper presents a novel Reinforcement Learning (RL) based approach to optimize robot inspection trajectories for profilometric sensors. Building upon the Boustrophedon scanning method, our technique dynamically adjusts the sensor position and tilt to maintain optimal orientation and distance from the surface, while also ensuring a consistent profile distance for uniform and high-quality scanning. Utilizing a simulated environment based on the CAD model of the part, we replicate real-world scanning conditions, including sensor noise and surface irregularities. This simulation-based approach enables offline trajectory planning based on CAD models. Key contributions include the modeling of the state space, action space, and reward function, specifically designed for inspection applications using profilometric sensors. We use Proximal Policy Optimization (PPO) algorithm to efficiently train the RL agent, demonstrating its capability to optimize inspection trajectories with profilometric sensors. To validate our approach, we conducted several experiments where a model trained on a specific training piece was tested on various parts in simulation. Also, we conducted a real-world experiment by executing the optimized trajectory, generated offline from a CAD model, to inspect a part using a UR3e robotic arm model.

This paper presents an online trajectory planning approach for optimal coordination of Fuel Cell (FC) and battery in plug-in Hybrid Electric Vehicle (HEV). One of the main challenges in energy management of plug-in HEV is generating State-of-Charge (SOC) reference curves by optimally depleting battery under high uncertainties in driving scenarios. Recent studies have begun to explore the potential of utilizing partial trip information for optimal SOC trajectory planning, but dynamic responses of the FC system are not taken into account. On the other hand, research focusing on dynamic operation of FC systems often focuses on air flow management, and battery has been treated only partially. Our aim is to fill this gap by designing an online trajectory planner for dynamic coordination of FC and battery systems that works with a high-level SOC planner in a hierarchical manner. We propose an iterative LQR based online trajectory planning method where the amount of electricity dischargeable at each driving segment can be explicitly and adaptively specified by the high-level planner. Numerical results are provided as a proof of concept example to show the effectiveness of the proposed approach.

Understanding human intentions and actions through egocentric videos is important on the path to embodied artificial intelligence. As a branch of egocentric vision techniques, hand trajectory prediction plays a vital role in comprehending human motion patterns, benefiting downstream tasks in extended reality and robot manipulation. However, capturing high-level human intentions consistent with reasonable temporal causality is challenging when only egocentric videos are available. This difficulty is exacerbated under camera egomotion interference and the absence of affordance labels to explicitly guide the optimization of hand waypoint distribution. In this work, we propose a novel hand trajectory prediction method dubbed MADiff, which forecasts future hand waypoints with diffusion models. The devised denoising operation in the latent space is achieved by our proposed motion-aware Mamba, where the camera wearer's egomotion is integrated to achieve motion-driven selective scan (MDSS). To discern the relationship between hands and scenarios without explicit affordance supervision, we leverage a foundation model that fuses visual and language features to capture high-level semantics from video clips. Comprehensive experiments conducted on five public datasets with the existing and our proposed new evaluation metrics demonstrate that MADiff predicts comparably reasonable hand trajectories compared to the state-of-the-art baselines, and achieves real-time performance. We will release our code and pretrained models of MADiff at the project page: https://irmvlab.github.io/madiff.github.io.

The estimated Glomerular Filtration Rate (eGFR) is an essential indicator of kidney function in clinical practice. Although traditional equations and Machine Learning (ML) models using clinical and laboratory data can estimate eGFR, accurately predicting future eGFR levels remains a significant challenge for nephrologists and ML researchers. Recent advances demonstrate that Large Language Models (LLMs) and Large Multimodal Models (LMMs) can serve as robust foundation models for diverse applications. This study investigates the potential of LMMs to predict future eGFR levels with a dataset consisting of laboratory and clinical values from 50 patients. By integrating various prompting techniques and ensembles of LMMs, our findings suggest that these models, when combined with precise prompts and visual representations of eGFR trajectories, offer predictive performance comparable to existing ML models. This research extends the application of foundation models and suggests avenues for future studies to harness these models in addressing complex medical forecasting challenges.

This preprint version includes corrections of typographical errors related to numerical values in Table 2, which were present in the version published at the BDH workshop in MIPR 2024. These corrections do not affect the overall conclusions of the study

Joint pedestrian trajectory prediction has long grappled with the inherent unpredictability of human behaviors. Recent investigations employing variants of conditional diffusion models in trajectory prediction have exhibited notable success. Nevertheless, the heavy dependence on accurate historical data results in their vulnerability to noise disturbances and data incompleteness. To improve the robustness and reliability, we introduce the Guided Full Trajectory Diffuser (GFTD), a novel diffusion model framework that captures the joint full (historical and future) trajectory distribution. By learning from the full trajectory, GFTD can recover the noisy and missing data, hence improving the robustness. In addition, GFTD can adapt to data imperfections without additional training requirements, leveraging posterior sampling for reliable prediction and controllable generation. Our approach not only simplifies the prediction process but also enhances generalizability in scenarios with noise and incomplete inputs. Through rigorous experimental evaluation, GFTD exhibits superior performance in both trajectory prediction and controllable generation.

This paper explores pedestrian trajectory prediction in urban traffic while focusing on both model accuracy and real-world applicability. While promising approaches exist, they are often not publicly available, revolve around pedestrian datasets excluding traffic-related information, or resemble architectures that are either not real-time capable or robust. To address these limitations, we first introduce a dedicated benchmark based on Argoverse 2, specifically targeting pedestrians in urban settings. Following this, we present Snapshot, a modular, feed-forward neural network that outperforms the current state of the art while utilizing significantly less information. Despite its agent-centric encoding scheme, Snapshot demonstrates scalability, real-time performance, and robustness to varying motion histories. Moreover, by integrating Snapshot into a modular autonomous driving software stack, we showcase its real-world applicability

Long-term trajectory forecasting is an important and challenging problem in the fields of computer vision, machine learning, and robotics. One fundamental difficulty stands in the evolution of the trajectory that becomes more and more uncertain and unpredictable as the time horizon grows, subsequently increasing the complexity of the problem. To overcome this issue, in this paper, we propose Di-Long, a new method that employs the distillation of a short-term trajectory model forecaster that guides a student network for long-term trajectory prediction during the training process. Given a total sequence length that comprehends the allowed observation for the student network and the complementary target sequence, we let the student and the teacher solve two different related tasks defined over the same full trajectory: the student observes a short sequence and predicts a long trajectory, whereas the teacher observes a longer sequence and predicts the remaining short target trajectory. The teacher's task is less uncertain, and we use its accurate predictions to guide the student through our knowledge distillation framework, reducing long-term future uncertainty. Our experiments show that our proposed Di-Long method is effective for long-term forecasting and achieves state-of-the-art performance on the Intersection Drone Dataset (inD) and the Stanford Drone Dataset (SDD).

Accepted to IROS 2024

This paper presents an innovative terminal landing guidance law that utilizes an analytic solution derived from the gravity turn trajectory. The characteristics of the derived solution are thoroughly investigated, and the solution is employed to generate a reference velocity vector that satisfies terminal landing conditions. A nonlinear control law is applied to effectively track the reference velocity vector within a finite time, and its robustness against disturbances is studied. Furthermore, the guidance law is expanded to incorporate ground collision avoidance by considering the shape of the gravity turn trajectory. The proposed method's fuel efficiency, robustness, and practicality are demonstrated through comprehensive numerical simulations, and its performance is compared with existing methods.

We propose semidefinite trajectory optimization (STROM), a framework that computes fast and certifiably optimal solutions for nonconvex trajectory optimization problems defined by polynomial objectives and constraints. STROM employs sparse second-order Lasserre's hierarchy to generate semidefinite program (SDP) relaxations of trajectory optimization. Different from existing tools (e.g., YALMIP and SOSTOOLS in Matlab), STROM generates chain-like multiple-block SDPs with only positive semidefinite (PSD) variables. Moreover, STROM does so two orders of magnitude faster. Underpinning STROM is cuADMM, the first ADMM-based SDP solver implemented in CUDA and runs in GPUs (with C/C++ extension). cuADMM builds upon the symmetric Gauss-Seidel ADMM algorithm and leverages GPU parallelization to speedup solving sparse linear systems and projecting onto PSD cones. In five trajectory optimization problems (inverted pendulum, cart-pole, vehicle landing, flying robot, and car back-in), cuADMM computes optimal trajectories (with certified suboptimality below 1%) in minutes (when other solvers take hours or run out of memory) and seconds (when others take minutes). Further, when warmstarted by data-driven initialization in the inverted pendulum problem, cuADMM delivers real-time performance: providing certifiably optimal trajectories in 0.66 seconds despite the SDP has 49,500 variables and 47,351 constraints.

We propose TRAM, a two-stage method to reconstruct a human's global trajectory and motion from in-the-wild videos. TRAM robustifies SLAM to recover the camera motion in the presence of dynamic humans and uses the scene background to derive the motion scale. Using the recovered camera as a metric-scale reference frame, we introduce a video transformer model (VIMO) to regress the kinematic body motion of a human. By composing the two motions, we achieve accurate recovery of 3D humans in the world space, reducing global motion errors by a large margin from prior work. https://yufu-wang.github.io/tram4d/

The project website: https://yufu-wang.github.io/tram4d/

Motion prediction plays an essential role in autonomous driving systems, enabling autonomous vehicles to achieve more accurate local-path planning and driving decisions based on predictions of the surrounding vehicles. However, existing methods neglect the potential missing values caused by object occlusion, perception failures, etc., which inevitably degrades the trajectory prediction performance in real traffic scenarios. To address this limitation, we propose a novel end-to-end framework for incomplete vehicle trajectory prediction, named Multi-scale Temporal Fusion Transformer (MTFT), which consists of the Multi-scale Attention Head (MAH) and the Continuity Representation-guided Multi-scale Fusion (CRMF) module. Specifically, the MAH leverages the multi-head attention mechanism to parallelly capture multi-scale motion representation of trajectory from different temporal granularities, thus mitigating the adverse effect of missing values on prediction. Furthermore, the multi-scale motion representation is input into the CRMF module for multi-scale fusion to obtain the robust temporal feature of the vehicle. During the fusion process, the continuity representation of vehicle motion is first extracted across time steps to guide the fusion, ensuring that the resulting temporal feature incorporates both detailed information and the overall trend of vehicle motion, which facilitates the accurate decoding of future trajectory that is consistent with the vehicle's motion trend. We evaluate the proposed model on four datasets derived from highway and urban traffic scenarios. The experimental results demonstrate its superior performance in the incomplete vehicle trajectory prediction task compared with state-of-the-art models, e.g., a comprehensive performance improvement of more than 39% on the HighD dataset.

Reinforcement learning (RL) trains an agent from experiences interacting with the environment. In scenarios where online interactions are impractical, offline RL, which trains the agent using pre-collected datasets, has become popular. While this new paradigm presents remarkable effectiveness across various real-world domains, like healthcare and energy management, there is a growing demand to enable agents to rapidly and completely eliminate the influence of specific trajectories from both the training dataset and the trained agents. To meet this problem, this paper advocates Trajdeleter, the first practical approach to trajectory unlearning for offline RL agents. The key idea of Trajdeleter is to guide the agent to demonstrate deteriorating performance when it encounters states associated with unlearning trajectories. Simultaneously, it ensures the agent maintains its original performance level when facing other remaining trajectories. Additionally, we introduce Trajauditor, a simple yet efficient method to evaluate whether Trajdeleter successfully eliminates the specific trajectories of influence from the offline RL agent. Extensive experiments conducted on six offline RL algorithms and three tasks demonstrate that Trajdeleter requires only about 1.5% of the time needed for retraining from scratch. It effectively unlearns an average of 94.8% of the targeted trajectories yet still performs well in actual environment interactions after unlearning. The replication package and agent parameters are available online.

Accepted at NDSS 2025. The presented document here is the full version of our paper

Patterns of human motion in outdoor and indoor environments are substantially different due to the scope of the environment and the typical intentions of people therein. While outdoor trajectory forecasting has received significant attention, indoor forecasting is still an underexplored research area. This paper proposes SITUATE, a novel approach to cope with indoor human trajectory prediction by leveraging equivariant and invariant geometric features and a self-supervised vision representation. The geometric learning modules model the intrinsic symmetries and human movements inherent in indoor spaces. This concept becomes particularly important because self-loops at various scales and rapid direction changes often characterize indoor trajectories. On the other hand, the vision representation module is used to acquire spatial-semantic information about the environment to predict users' future locations more accurately. We evaluate our method through comprehensive experiments on the two most famous indoor trajectory forecasting datasets, i.e., TH\"OR and Supermarket, obtaining state-of-the-art performance. Furthermore, we also achieve competitive results in outdoor scenarios, showing that indoor-oriented forecasting models generalize better than outdoor-oriented ones. The source code is available at https://github.com/intelligolabs/SITUATE.

Accepted at the 27th International Conference on Pattern Recognition (ICPR 2024)

With the proliferation of location-aware devices, large amount of trajectories have been generated when agents such as people, vehicles and goods flow around the urban environment. These raw trajectories, typically collected from various sources such as GPS in cars, personal mobile devices, and public transport, are often sparse and fragmented due to limited sampling rates, infrastructure coverage and data loss. In this context, trajectory recovery aims to reconstruct such sparse raw trajectories into their dense and continuous counterparts, so that fine-grained movement of agents across space and time can be captured faithfully. Existing trajectory recovery approaches typically rely on the prior knowledge of travel mode or motion patterns, and often fail in densely populated urban areas where accurate maps are absent. In this paper, we present a new recovery framework called TrajWeaver based on probabilistic diffusion models, which is able to recover dense and refined trajectories from the sparse raw ones, conditioned on various auxiliary features such as Areas of Interest along the way, user identity and waybill information. The core of TrajWeaver is a novel State Propagation Diffusion Model (SPDM), which introduces a new state propagation mechanism on top of the standard diffusion models, so that knowledge computed in earlier diffusion steps can be reused later, improving the recovery performance while reducing the number of steps needed. Extensive experiments show that the proposed TrajWeaver can recover from raw trajectories of various lengths, sparsity levels and heterogeneous travel modes, and outperform the state-of-the-art baselines significantly in recovery accuracy. Our code is available at: https://anonymous.4open.science/r/TrajWeaver/

First submission, extended to 10 pages include ref

Traffic roundabouts, as complex and critical road scenarios, pose significant safety challenges for autonomous vehicles. In particular, the encounter of a vehicle with a dilemma zone (DZ) at a roundabout intersection is a pivotal concern. This paper presents an automated system that leverages trajectory forecasting to predict DZ events, specifically at traffic roundabouts. Our system aims to enhance safety standards in both autonomous and manual transportation. The core of our approach is a modular, graph-structured recurrent model that forecasts the trajectories of diverse agents, taking into account agent dynamics and integrating heterogeneous data, such as semantic maps. This model, based on graph neural networks, aids in predicting DZ events and enhances traffic management decision-making. We evaluated our system using a real-world dataset of traffic roundabout intersections. Our experimental results demonstrate that our dilemma forecasting system achieves a high precision with a low false positive rate of 0.1. This research represents an advancement in roundabout DZ data mining and forecasting, contributing to the assurance of intersection safety in the era of autonomous vehicles.

This research focuses on assessing the ability of AI foundation models in representing the trajectories of movements. We utilize one of the large language models (LLMs) (i.e., GPT-J) to encode the string format of trajectories and then evaluate the effectiveness of the LLM-based representation for trajectory data analysis. The experiments demonstrate that while the LLM-based embeddings can preserve certain trajectory distance metrics (i.e., the correlation coefficients exceed 0.74 between the Cosine distance derived from GPT-J embeddings and the Hausdorff and Dynamic Time Warping distances on raw trajectories), challenges remain in restoring numeric values and retrieving spatial neighbors in movement trajectory analytics. In addition, the LLMs can understand the spatiotemporal dependency contained in trajectories and have good accuracy in location prediction tasks. This research highlights the need for improvement in terms of capturing the nuances and complexities of the underlying geospatial data and integrating domain knowledge to support various GeoAI applications using LLMs.

Performing trajectory design for humanoid robots with high degrees of freedom is computationally challenging. The trajectory design process also often involves carefully selecting various hyperparameters and requires a good initial guess which can further complicate the development process. This work introduces a generalized gait optimization framework that directly generates smooth and physically feasible trajectories. The proposed method demonstrates faster and more robust convergence than existing techniques and explicitly incorporates closed-loop kinematic constraints that appear in many modern humanoids. The method is implemented as an open-source C++ codebase which can be found at https://roahmlab.github.io/RAPTOR/.

We present a novel end-to-end diffusion-based trajectory generation method, DTG, for mapless global navigation in challenging outdoor scenarios with occlusions and unstructured off-road features like grass, buildings, bushes, etc. Given a distant goal, our approach computes a trajectory that satisfies the following goals: (1) minimize the travel distance to the goal; (2) maximize the traversability by choosing paths that do not lie in undesirable areas. Specifically, we present a novel Conditional RNN(CRNN) for diffusion models to efficiently generate trajectories. Furthermore, we propose an adaptive training method that ensures that the diffusion model generates more traversable trajectories. We evaluate our methods in various outdoor scenes and compare the performance with other global navigation algorithms on a Husky robot. In practice, we observe at least a 15% improvement in traveling distance and around a 7% improvement in traversability.

In Model Predictive Control (MPC), discrepancies between the actual system and the predictive model can lead to substantial tracking errors and significantly degrade performance and reliability. While such discrepancies can be alleviated with more complex models, this often complicates controller design and implementation. By leveraging the fact that many trajectories of interest are periodic, we show that perfect tracking is possible when incorporating a simple observer that estimates and compensates for periodic disturbances. We present the design of the observer and the accompanying tracking MPC scheme, proving that their combination achieves zero tracking error asymptotically, regardless of the complexity of the unmodelled dynamics. We validate the effectiveness of our method, demonstrating asymptotically perfect tracking on a high-dimensional soft robot with nearly 10,000 states and a fivefold reduction in tracking errors compared to a baseline MPC on small-scale autonomous race car experiments.

8 pages, 3 figures; 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)

The resurgence of lunar operations requires advancements in cislunar navigation and Space Situational Awareness (SSA). Challenges associated to these tasks have created an interest in autonomous planning, navigation, and tracking technologies that operate with little ground-based intervention. This research introduces a trajectory planning tool for a low-thrust mobile observer, aimed at maximizing navigation and tracking performance with satellite-to-satellite relative measurements. We formulate an expression for the information gathered over an observation period based on the mutual information between augmented observer/target states and the associated measurement set collected. We then develop an optimal trajectory design problem for a mobile observer, balancing information gain and control effort, and solve this problem with a Sequential Convex Programming (SCP) approach. The developed methods are demonstrated in scenarios involving spacecraft in the cislunar regime, demonstrating the potential for improved autonomous navigation and tracking.

2024 AAS/AIAA Astrodynamics Specialist Conference

Model-based reinforcement learning (RL) is anticipated to exhibit higher sample efficiency compared to model-free RL by utilizing a virtual environment model. However, it is challenging to obtain sufficiently accurate representations of the environmental dynamics due to uncertainties in complex systems and environments. An inaccurate environment model may degrade the sample efficiency and performance of model-based RL. Furthermore, while model-based RL can improve sample efficiency, it often still requires substantial training time to learn from scratch, potentially limiting its advantages over model-free approaches. To address these challenges, this paper introduces a knowledge-informed model-based residual reinforcement learning framework aimed at enhancing learning efficiency by infusing established expert knowledge into the learning process and avoiding the issue of beginning from zero. Our approach integrates traffic expert knowledge into a virtual environment model, employing the Intelligent Driver Model (IDM) for basic dynamics and neural networks for residual dynamics, thus ensuring adaptability to complex scenarios. We propose a novel strategy that combines traditional control methods with residual RL, facilitating efficient learning and policy optimization without the need to learn from scratch. The proposed approach is applied to CAV trajectory control tasks for the dissipation of stop-and-go waves in mixed traffic flow. Experimental results demonstrate that our proposed approach enables the CAV agent to achieve superior performance in trajectory control compared to the baseline agents in terms of sample efficiency, traffic flow smoothness and traffic mobility. The source code and supplementary materials are available at https://github.com/zihaosheng/traffic-expertise-RL/.

In this paper, a fixed-time disturbance observerbased model predictive control algorithm is proposed for trajectory tracking of quadrotor in the presence of disturbances. First, a novel multivariable fixed-time disturbance observer is proposed to estimate the lumped disturbances. The bi-limit homogeneity and Lyapunov techniques are employed to ensure the convergence of estimation error within a fixed convergence time, independent of the initial estimation error. Then, an observerbased model predictive control strategy is formulated to achieve robust trajectory tracking of quadrotor, attenuating the lumped disturbances and model uncertainties. Finally, simulations and real-world experiments are provided to illustrate the effectiveness of the proposed method.

Trajectory forecasting is crucial for video surveillance analytics, as it enables the anticipation of future movements for a set of agents, e.g. basketball players engaged in intricate interactions with long-term intentions. Deep generative models offer a natural learning approach for trajectory forecasting, yet they encounter difficulties in achieving an optimal balance between sampling fidelity and diversity. We address this challenge by leveraging Vector Quantized Variational Autoencoders (VQ-VAEs), which utilize a discrete latent space to tackle the issue of posterior collapse. Specifically, we introduce an instance-based codebook that allows tailored latent representations for each example. In a nutshell, the rows of the codebook are dynamically adjusted to reflect contextual information (i.e., past motion patterns extracted from the observed trajectories). In this way, the discretization process gains flexibility, leading to improved reconstructions. Notably, instance-level dynamics are injected into the codebook through low-rank updates, which restrict the customization of the codebook to a lower dimension space. The resulting discrete space serves as the basis of the subsequent step, which regards the training of a diffusion-based predictive model. We show that such a two-fold framework, augmented with instance-level discretization, leads to accurate and diverse forecasts, yielding state-of-the-art performance on three established benchmarks.

15 pages, 3 figures, 5 tables

To plan the trajectories of a large-scale heterogeneous swarm, sequentially or synchronously distributed methods usually become intractable due to the lack of global clock synchronization. To this end, we provide a novel asynchronous spatial-temporal allocation method. Specifically, between a pair of agents, the allocation is proposed to determine their corresponding derivable time-stamped space and can be updated in an asynchronous way, by inserting a waiting duration between two consecutive replanning steps. Via theoretical analysis, the inter-agent collision is proved to be avoided and the allocation ensures timely updates. Comprehensive simulations and comparisons with five baselines validate the effectiveness of the proposed method and illustrate its improvement in completion time and moving distance. Finally, hardware experiments are carried out, where $8$ heterogeneous unmanned ground vehicles with onboard computation navigate in cluttered scenarios with high agility.

Non-prehensile object transportation offers a way to enhance robotic performance in object manipulation tasks, especially with unstable objects. Effective trajectory planning requires simultaneous consideration of robot motion constraints and object stability. Here, we introduce a physical model for object stability and propose a novel trajectory planning approach for non-prehensile transportation along arbitrary straight lines in 3D space. Validation with a 7-DoF Franka Panda robot confirms improved transportation speed via tray rotation integration while ensuring object stability and robot motion constraints.

Accepted to the European Robotic Forum (ERF) 2024

Accumulating substantial volumes of real-world driving data proves pivotal in the realm of trajectory forecasting for autonomous driving. Given the heavy reliance of current trajectory forecasting models on data-driven methodologies, we aim to tackle the challenge of learning general trajectory forecasting representations under limited data availability. We propose a pipeline-level solution to mitigate the issue of data scarcity in trajectory forecasting. The solution is composed of two parts: firstly, we adopt HD map augmentation and trajectory synthesis for generating driving data, and then we learn representations by pre-training on them. Specifically, we apply vector transformations to reshape the maps, and then employ a rule-based model to generate trajectories on both original and augmented scenes; thus enlarging the driving data without collecting additional real ones. To foster the learning of general representations within this augmented dataset, we comprehensively explore the different pre-training strategies, including extending the concept of a Masked AutoEncoder (MAE) for trajectory forecasting. Without bells and whistles, our proposed pipeline-level solution is general, simple, yet effective: we conduct extensive experiments to demonstrate the effectiveness of our data expansion and pre-training strategies, which outperform the baseline prediction model by large margins, e.g. 5.04%, 3.84% and 8.30% in terms of $MR_6$, $minADE_6$ and $minFDE_6$. The pre-training dataset and the codes for pre-training and fine-tuning are released at https://github.com/yhli123/Pretraining_on_Synthetic_Driving_Data_for_Trajectory_Prediction.

Accelerating the sampling speed of diffusion models remains a significant challenge. Recent score distillation methods distill a heavy teacher model into an one-step student generator, which is optimized by calculating the difference between the two score functions on the samples generated by the student model. However, there is a score mismatch issue in the early stage of the distillation process, because existing methods mainly focus on using the endpoint of pre-trained diffusion models as teacher models, overlooking the importance of the convergence trajectory between the student generator and the teacher model. To address this issue, we extend the score distillation process by introducing the entire convergence trajectory of teacher models and propose Distribution Backtracking Distillation (DisBack) for distilling student generators. DisBask is composed of two stages: Degradation Recording and Distribution Backtracking. Degradation Recording is designed to obtain the convergence trajectory of teacher models, which records the degradation path from the trained teacher model to the untrained initial student generator. The degradation path implicitly represents the intermediate distributions of teacher models. Then Distribution Backtracking trains a student generator to backtrack the intermediate distributions for approximating the convergence trajectory of teacher models. Extensive experiments show that DisBack achieves faster and better convergence than the existing distillation method and accomplishes comparable generation performance. Notably, DisBack is easy to implement and can be generalized to existing distillation methods to boost performance. Our code is publicly available on https://github.com/SYZhang0805/DisBack.

The problem of career trajectory prediction (CTP) aims to predict one's future employer or job position. While several CTP methods have been developed for this problem, we posit that none of these methods (1) jointly considers the mutual ternary dependency between three key units (i.e., user, position, and company) of a career and (2) captures the characteristic shifts of key units in career over time, leading to an inaccurate understanding of the job movement patterns in the labor market. To address the above challenges, we propose a novel solution, named as CAPER, that solves the challenges via sophisticated temporal knowledge graph (TKG) modeling. It enables the utilization of a graph-structured knowledge base with rich expressiveness, effectively preserving the changes in job movement patterns. Furthermore, we devise an extrapolated career reasoning task on TKG for a realistic evaluation. The experiments on a real-world career trajectory dataset demonstrate that CAPER consistently and significantly outperforms four baselines, two recent TKG reasoning methods, and five state-of-the-art CTP methods in predicting one's future companies and positions-i.e., on average, yielding 6.80% and 34.58% more accurate predictions, respectively.

For extending CT field-of-view to perform non-destructive testing, the Symmetric Multi-Linear trajectory Computed Tomography (SMLCT) has been developed as a successful example of non-standard CT scanning modes. However, inevitable geometric errors can cause severe artifacts in the reconstructed images. The existing calibration method for SMLCT is both crude and inefficient. It involves reconstructing hundreds of images by exhaustively substituting each potential error, and then manually identifying the images with the fewest geometric artifacts to estimate the final geometric errors for calibration. In this paper, we comprehensively and efficiently address the challenging geometric artifacts in SMLCT, , and the corresponding works mainly involve theory, method, and generalization. In particular, after identifying sensitive parameters and conducting some theory analysis of geometric artifacts, we summarize several key properties between sensitive geometric parameters and artifact characteristics. Then, we further construct mathematical relationships that relate sensitive geometric errors to the pixel offsets of reconstruction images with artifact characteristics. To accurately extract pixel bias, we innovatively adapt the Generalized Cross-Correlation with Phase Transform (GCC-PHAT) algorithm, commonly used in sound processing, for our image registration task for each paired symmetric LCT. This adaptation leads to the design of a highly efficient rigid translation registration method. Simulation and physical experiments have validated the excellent performance of this work. Additionally, our results demonstrate significant generalization to common rotated CT and a variant of SMLCT.

Target tracking has numerous significant civilian and military applications, and maintaining the visibility of the target plays a vital role in ensuring the success of the tracking task. Existing visibility-aware planners primarily focus on keeping the target within the limited field of view of an onboard sensor and avoiding obstacle occlusion. However, the negative impact of system uncertainty is often neglected, rendering the planners delicate to uncertainties in practice. To bridge the gap, this work proposes a real-time, non-myopic trajectory planner for visibility-aware and safe target tracking in the presence of system uncertainty. For more accurate target motion prediction, we introduce the concept of belief-space probability of detection (BPOD) to measure the predictive visibility of the target under stochastic robot and target states. An Extended Kalman Filter variant incorporating BPOD is developed to predict target belief state under uncertain visibility within the planning horizon. To reach real-time trajectory planning, we propose a computationally efficient algorithm to uniformly calculate both BPOD and the chance-constrained collision risk by utilizing linearized signed distance function (SDF), and then design a two-stage strategy for lightweight calculation of SDF in sequential convex programming. Extensive simulation results with benchmark comparisons show the capacity of the proposed approach to robustly maintain the visibility of the target under high system uncertainty. The practicality of the proposed trajectory planner is validated by real-world experiments.

A technical report for our conference paper in 2024 American Control Conference (ACC)

Recent advancements in Diffusion Transformer (DiT) have demonstrated remarkable proficiency in producing high-quality video content. Nonetheless, the potential of transformer-based diffusion models for effectively generating videos with controllable motion remains an area of limited exploration. This paper introduces Tora, the first trajectory-oriented DiT framework that concurrently integrates textual, visual, and trajectory conditions, thereby enabling scalable video generation with effective motion guidance. Specifically, Tora consists of a Trajectory Extractor(TE), a Spatial-Temporal DiT, and a Motion-guidance Fuser(MGF). The TE encodes arbitrary trajectories into hierarchical spacetime motion patches with a 3D video compression network. The MGF integrates the motion patches into the DiT blocks to generate consistent videos that accurately follow designated trajectories. Our design aligns seamlessly with DiT's scalability, allowing precise control of video content's dynamics with diverse durations, aspect ratios, and resolutions. Extensive experiments demonstrate Tora's excellence in achieving high motion fidelity, while also meticulously simulating the intricate movement of the physical world.

Despite the widespread adoption of multi-task training in deep learning, little is understood about how multi-task learning (MTL) affects generalization. Prior work has conjectured that the negative effects of MTL are due to optimization challenges that arise during training, and many optimization methods have been proposed to improve multi-task performance. However, recent work has shown that these methods fail to consistently improve multi-task generalization. In this work, we seek to improve our understanding of these failures by empirically studying how MTL impacts the optimization of tasks, and whether this impact can explain the effects of MTL on generalization. We show that MTL results in a generalization gap-a gap in generalization at comparable training loss-between single-task and multi-task trajectories early into training. However, we find that factors of the optimization trajectory previously proposed to explain generalization gaps in single-task settings cannot explain the generalization gaps between single-task and multi-task models. Moreover, we show that the amount of gradient conflict between tasks is correlated with negative effects to task optimization, but is not predictive of generalization. Our work sheds light on the underlying causes for failures in MTL and, importantly, raises questions about the role of general purpose multi-task optimization algorithms.

In this work, we propose a trajectory optimization approach for robot navigation in cluttered 3D environments. We represent the robot's geometry as a semialgebraic set defined by polynomial inequalities such that robots with general shapes can be suitably characterized. To address the robot navigation task in obstacle-dense environments, we exploit the free space directly to construct a sequence of free regions, and allocate each waypoint on the trajectory to a specific region. Then, we incorporate a uniform scaling factor for each free region, and formulate a Sums-of-Squares (SOS) optimization problem that renders the containment relationship between the robot and the free space computationally tractable. The SOS optimization problem is further reformulated to a semidefinite program (SDP), and the collision-free constraints are shown to be equivalent to limiting the scaling factor along the entire trajectory. In this context, the robot at a specific configuration is tailored to stay within the free region. Next, to solve the trajectory optimization problem with the proposed safety constraints (which are implicitly dependent on the robot configurations), we derive the analytical solution to the gradient of the minimum scaling factor with respect to the robot configuration. As a result, this seamlessly facilitates the use of gradient-based methods in efficient solving of the trajectory optimization problem. Through a series of simulations and real-world experiments, the proposed trajectory optimization approach is validated in various challenging scenarios, and the results demonstrate its effectiveness in generating collision-free trajectories in dense and intricate environments populated with obstacles. Our code is available at: https://github.com/lyl00/minimum_scaling_free_region

Robustness against Out-of-Distribution (OoD) samples is a key performance indicator of a trajectory prediction model. However, the development and ranking of state-of-the-art (SotA) models are driven by their In-Distribution (ID) performance on individual competition datasets. We present an OoD testing protocol that homogenizes datasets and prediction tasks across two large-scale motion datasets. We introduce a novel prediction algorithm based on polynomial representations for agent trajectory and road geometry on both the input and output sides of the model. With a much smaller model size, training effort, and inference time, we reach near SotA performance for ID testing and significantly improve robustness in OoD testing. Within our OoD testing protocol, we further study two augmentation strategies of SotA models and their effects on model generalization. Highlighting the contrast between ID and OoD performance, we suggest adding OoD testing to the evaluation criteria of trajectory prediction models.

This paper addresses motion forecasting in multi-agent environments, pivotal for ensuring safety of autonomous vehicles. Traditional as well as recent data-driven marginal trajectory prediction methods struggle to properly learn non-linear agent-to-agent interactions. We present SSL-Interactions that proposes pretext tasks to enhance interaction modeling for trajectory prediction. We introduce four interaction-aware pretext tasks to encapsulate various aspects of agent interactions: range gap prediction, closest distance prediction, direction of movement prediction, and type of interaction prediction. We further propose an approach to curate interaction-heavy scenarios from datasets. This curated data has two advantages: it provides a stronger learning signal to the interaction model, and facilitates generation of pseudo-labels for interaction-centric pretext tasks. We also propose three new metrics specifically designed to evaluate predictions in interactive scenes. Our empirical evaluations indicate SSL-Interactions outperforms state-of-the-art motion forecasting methods quantitatively with up to 8% improvement, and qualitatively, for interaction-heavy scenarios.

Accepted at IV-2024. 13 pages, 5 figures

Recent advancements in Large Language Models (LLMs) have led to significant breakthroughs in various natural language processing tasks. However, generating factually consistent responses in knowledge-intensive scenarios remains a challenge due to issues such as hallucination, difficulty in acquiring long-tailed knowledge, and limited memory expansion. This paper introduces SMART, a novel multi-agent framework that leverages external knowledge to enhance the interpretability and factual consistency of LLM-generated responses. SMART comprises four specialized agents, each performing a specific sub-trajectory action to navigate complex knowledge-intensive tasks. We propose a multi-agent co-training paradigm, Long-Short Trajectory Learning, which ensures synergistic collaboration among agents while maintaining fine-grained execution by each agent. Extensive experiments on five knowledge-intensive tasks demonstrate SMART's superior performance compared to widely adopted knowledge internalization and knowledge enhancement methods. Our framework can extend beyond knowledge-intensive tasks to more complex scenarios. Our code is available at https://github.com/yueshengbin/SMART.

We propose a global constructibility analysis for a vehicle moving on a planar surface. Assuming that the vehicle follows a trajectory that can be uniquely identified by the sequence of control inputs and by some intermittent ranging measurements from known points in the environment, we can model the trajectory as a rigid body subject to rotation and translation in the plane. This way, the localisation problem can be reduced to finding the conditions for the existence of a unique roto-translation of the trajectory from a known reference frame to the world reference frame, given the collected measurements. As discussed in this paper, such conditions can be expressed in terms of the shape of the trajectory, of the layout of the ranging sensors, and of the numbers of measurements collected from each of them. The approach applies to a large class of kinematic models. Focusing on the special case of unicycle kinematics, we provide additional local constructibility results.

12 pages, 7 figures, This article has been accepted for publication in IEEE Transactions on Automatic Control (2024). content may change prior to final publication

Parameterizing finger rolling and finger-object contacts in a differentiable manner is important for formulating dexterous manipulation as a trajectory optimization problem. In contrast to previous methods which often assume simplified geometries of the robot and object or do not explicitly model finger rolling, we propose a method to further extend the capabilities of dexterous manipulation by accounting for non-trivial geometries of both the robot and the object. By integrating the object's Signed Distance Field (SDF) with a sampling method, our method estimates contact and rolling-related variables and includes those in a trajectory optimization framework. This formulation naturally allows for the emergence of finger-rolling behaviors, enabling the robot to locally adjust the contact points. Our method is tested in a peg alignment task and a screwdriver turning task, where it outperforms the baselines in terms of achieving desired object configurations and avoiding dropping the object. We also successfully apply our method to a real-world screwdriver turning task, demonstrating its robustness to the sim2real gap.

The focus of this paper is a key component of a methodology for understanding, interpolating, and predicting fish movement patterns based on spatiotemporal data recorded by spatially static acoustic receivers. For periods of time, fish may be far from the receivers, resulting in the absence of observations. The lack of information on the fish's location for extended time periods poses challenges to the understanding of fish movement patterns, and hence, the identification of proper statistical inference frameworks for modeling the trajectories. As the initial step in our methodology, in this paper, we implement an imputation strategy that relies on both Markov chain and Brownian motion principles to enhance our dataset over time. This methodology will be generalizable and applicable to all fish species with similar migration patterns or data with similar structures due to the use of static acoustic receivers.

Ship trajectories from Automatic Identification System (AIS) messages are important in maritime safety, domain awareness, and algorithmic testing. Although the specifications for transmitting and receiving AIS messages are fixed, it is well known that technical inaccuracies and lacking seafarer compliance lead to severe data quality impairment. This paper proposes an adaptable, data-driven, maneuverability-dependent, $\alpha$-quantile-based framework for decoding, constructing, splitting, and assessing trajectories from raw AIS records to improve transparency in AIS data mining. Results indicate the proposed filtering algorithm robustly extracts clean, long, and uninterrupted trajectories for further processing. An open-source Python implementation of the framework is provided.

Spatiotemporal trajectories collected from GPS-enabled devices are of vital importance to many applications, such as urban planning and traffic analysis. Due to the privacy leakage concerns, many privacy-preserving trajectory publishing methods have been proposed. However, most of them could not strike a good balance between privacy protection and good data utility. In this paper, we propose DP-STTS, a differentially private spatiotemporal trajectory synthesizer with high data utility, which employs a model composed of a start spatiotemporal cube distribution and a 1-order Markov process. Specially, DP-STTS firstly discretizes the raw spatiotemporal trajectories into neighboring cubes, such that the model size is limited and the model's tolerance for noise could be enhanced. Then, a Markov process is utilized for the next location point picking. After adding noise under differential privacy (DP) to the model, synthetic trajectories that preserve essential spatial and temporal characteristics of the real trajectories are generated from the noisy model. Experiments on one real-life dataset demonstrate that DP-STTS provides good data utility. Our code is available at https://github.com/Etherious72/DP-STTS.

Movable antennas (MAs) have received increasing attention in wireless communications due to their capability of antenna position adjustment to reconfigure wireless channels. However, moving MAs results in non-negligible delay, which may decrease the effective data transmission time. To reduce the movement delay, we study in this paper a new MA trajectory optimization problem. In particular, given the desired destination positions of multiple MAs, we aim to jointly optimize their associations with the initial MA positions and the trajectories for moving them from their respective initial to destination positions within a given two-dimensional (2D) region, such that the delay of antenna movement is minimized, subject to the inter-MA minimum distance constraints in the movement. However, this problem is a continuous-time mixed-integer linear programming (MILP) problem that is challenging to solve. To tackle this challenge, we propose a two-stage optimization framework that sequentially optimizes the MAs' position associations and trajectories, respectively. First, we relax the inter-MA distance constraints and optimally solve the resulted delay minimization problem. Next, we check if the obtained MA association and trajectory solutions satisfy the inter-MA distance constraints. If not satisfied, we then employ a successive convex approximation (SCA) algorithm to adjust the MAs' trajectories until they satisfy the given constraints. Simulation results are provided to show the effectiveness of our proposed trajectory optimization method in reducing the movement delay as well as draw useful insights.

6 pages,6 figures, submit to GLOBECOM 2024 Workshop - IRAFWCC

Many spatiotemporal domains handle multi-agent trajectory data, but in real-world scenarios, collected trajectory data are often partially missing due to various reasons. While existing approaches demonstrate good performance in trajectory imputation, they face challenges in capturing the complex dynamics and interactions between agents due to a lack of physical constraints that govern realistic trajectories, leading to suboptimal results. To address this issue, the paper proposes a Derivative-Based Hybrid Prediction (DBHP) framework that can effectively impute multiple agents' missing trajectories. First, a neural network equipped with Set Transformers produces a naive prediction of missing trajectories while satisfying the permutation-equivariance in terms of the order of input agents. Then, the framework makes alternative predictions leveraging velocity and acceleration information and combines all the predictions with properly determined weights to provide final imputed trajectories. In this way, our proposed framework not only accurately predicts position, velocity, and acceleration values but also enforces the physical relationship between them, eventually improving both the accuracy and naturalness of the predicted trajectories. Accordingly, the experiment results about imputing player trajectories in team sports show that our framework significantly outperforms existing imputation baselines.

Gathering visual information effectively to monitor known environments is a key challenge in robotics. To be as efficient as human surveyors, robotic systems must continuously collect observational data required to complete their survey task. Inspection personnel instinctively know to look at relevant equipment that happens to be ``along the way.'' In this paper, we introduce a novel framework for continuous long-horizon viewpoint planning, for ground robots, applied to tasks involving patrolling, monitoring or visual data gathering in known environments. Our approach to Long Horizon Viewpoint Planning (LHVP), enables the robot to autonomously navigate and collect environmental data optimizing for coverage over the horizon of the patrol. Leveraging a quadruped's mobility and sensory capabilities, our LHVP framework plans patrol paths that account for coupling the viewpoint planner for the arm camera with the mobile base's navigation planner. The viewpath optimization algorithm seeks a balance between comprehensive environmental coverage and dynamically feasible movements, thus ensuring prolonged and effective operation in scenarios including monitoring, security surveillance, and disaster response. We validate our approach through simulations and in the real world and show that our LHVP significantly outperforms naive patrolling methods in terms of area coverage generating information-gathering trajectories for the robot arm. Our results indicate a promising direction for the deployment of mobile robots in long-term, autonomous surveying, and environmental data collection tasks, highlighting the potential of intelligent robotic systems in challenging real-world applications.

7 pages, 8 figures, 5 tables

High-level robot skills represent an increasingly popular paradigm in robot programming. However, configuring the skills' parameters for a specific task remains a manual and time-consuming endeavor. Existing approaches for learning or optimizing these parameters often require numerous real-world executions or do not work in dynamic environments. To address these challenges, we propose MuTT, a novel encoder-decoder transformer architecture designed to predict environment-aware executions of robot skills by integrating vision, trajectory, and robot skill parameters. Notably, we pioneer the fusion of vision and trajectory, introducing a novel trajectory projection. Furthermore, we illustrate MuTT's efficacy as a predictor when combined with a model-based robot skill optimizer. This approach facilitates the optimization of robot skill parameters for the current environment, without the need for real-world executions during optimization. Designed for compatibility with any representation of robot skills, MuTT demonstrates its versatility across three comprehensive experiments, showcasing superior performance across two different skill representations.

Trajectory tracking for an Omni-drive robot presents a challenging task that demands an efficient controller design. This paper introduces a self-optimizing controller, Type-1 fuzzyPID, which leverages dynamic and static system response analysis to overcome the limitations of manual tuning. To account for system uncertainties, an Interval Type-2 fuzzyPID controller is also developed. Both controllers are designed using Matlab/Simulink and tested through trajectory tracking simulations in the CoppeliaSim environment. Additionally, a non-linear model predictive controller(NMPC) is proposed and compared against the fuzzyPID controllers. The impact of tunable parameters on NMPC tracking accuracy is thoroughly examined. We also present plots of the step-response characteristics and noise rejection experiments for each controller. Simulation results validate the precision and effectiveness of NMPC over fuzzyPID controllers while trading computational complexity. Access to code and simulation environment is available in the following link: https://github.com/love481/Omni-drive-robot-Simulation.git.

As Chile's electric power sector advances toward a future powered by renewable energy, accurate forecasting of renewable generation is essential for managing grid operations. The integration of renewable energy sources is particularly challenging due to the operational difficulties of managing their power generation, which is highly variable compared to fossil fuel sources, delaying the availability of clean energy. To mitigate this, we quantify the impact of increasing intermittent generation from wind and solar on thermal power plants in Chile and introduce a hybrid wind speed forecasting methodology which combines two custom ML models for Chile. The first model is based on TiDE, an MLP-based ML model for short-term forecasts, and the second is based on a graph neural network, GraphCast, for medium-term forecasts up to 10 days. Our hybrid approach outperforms the most accurate operational deterministic systems by 4-21% for short-term forecasts and 5-23% for medium-term forecasts and can directly lower the impact of wind generation on thermal ramping, curtailment, and system-level emissions in Chile.

Graph Neural Networks based on the message-passing (MP) mechanism are a dominant approach for handling graph-structured data. However, they are inherently limited to modeling only pairwise interactions, making it difficult to explicitly capture the complexity of systems with $n$-body relations. To address this, topological deep learning has emerged as a promising field for studying and modeling higher-order interactions using various topological domains, such as simplicial and cellular complexes. While these new domains provide powerful representations, they introduce new challenges, such as effectively modeling the interactions among higher-order structures through higher-order MP. Meanwhile, structured state-space sequence models have proven to be effective for sequence modeling and have recently been adapted for graph data by encoding the neighborhood of a node as a sequence, thereby avoiding the MP mechanism. In this work, we propose a novel architecture designed to operate with simplicial complexes, utilizing the Mamba state-space model as its backbone. Our approach generates sequences for the nodes based on the neighboring cells, enabling direct communication between all higher-order structures, regardless of their rank. We extensively validate our model, demonstrating that it achieves competitive performance compared to state-of-the-art models developed for simplicial complexes.

Understanding the relationships between geometric structures and semantic concepts is crucial for building accurate models of complex environments. In indoors, certain spatial constraints, such as the relative positioning of planes, remain consistent despite variations in layout. This paper explores how these invariant relationships can be captured in a graph SLAM framework by representing high-level concepts like rooms and walls, linking them to geometric elements like planes through an optimizable factor graph. Several efforts have tackled this issue with add-hoc solutions for each concept generation and with manually-defined factors. This paper proposes a novel method for metric-semantic factor graph generation which includes defining a semantic scene graph, integrating geometric information, and learning the interconnecting factors, all based on Graph Neural Networks (GNNs). An edge classification network (G-GNN) sorts the edges between planes into same room, same wall or none types. The resulting relations are clustered, generating a room or wall for each cluster. A second family of networks (F-GNN) infers the geometrical origin of the new nodes. The definition of the factors employs the same F-GNN used for the metric attribute of the generated nodes. Furthermore, share the new factor graph with the S-Graphs+ algorithm, extending its graph expressiveness and scene representation with the ultimate goal of improving the SLAM performance. The complexity of the environments is increased to N-plane rooms by training the networks on L-shaped rooms. The framework is evaluated in synthetic and simulated scenarios as no real datasets of the required complex layouts are available.

Submitted to ICRA 2025

Advancement in finite element methods have become essential in various disciplines, and in particular for Computational Fluid Dynamics (CFD), driving research efforts for improved precision and efficiency. While Convolutional Neural Networks (CNNs) have found success in CFD by mapping meshes into images, recent attention has turned to leveraging Graph Neural Networks (GNNs) for direct mesh processing. This paper introduces a novel model merging Self-Attention with Message Passing in GNNs, achieving a 15\% reduction in RMSE on the well known flow past a cylinder benchmark. Furthermore, a dynamic mesh pruning technique based on Self-Attention is proposed, that leads to a robust GNN-based multigrid approach, also reducing RMSE by 15\%. Additionally, a new self-supervised training method based on BERT is presented, resulting in a 25\% RMSE reduction. The paper includes an ablation study and outperforms state-of-the-art models on several challenging datasets, promising advancements similar to those recently achieved in natural language and image processing. Finally, the paper introduces a dataset with meshes larger than existing ones by at least an order of magnitude. Code and Datasets will be released at https://github.com/DonsetPG/multigrid-gnn.

Graph-structured data naturally occurs in many research fields, such as chemistry and sociology. The relational information contained therein can be leveraged to statistically model graph properties through geometrical deep learning. Graph neural networks employ techniques, such as message-passing layers, to propagate local features through a graph. However, message-passing layers can be computationally expensive when dealing with large and sparse graphs. Graph pooling operators offer the possibility of removing or merging nodes in such graphs, thus lowering computational costs. However, pooling operators that remove nodes cause data loss, and pooling operators that merge nodes are often computationally expensive. We propose a pooling operator that merges nodes so as not to cause data loss but is also conceptually simple and computationally inexpensive. We empirically demonstrate that the proposed pooling operator performs statistically significantly better than edge pool on four popular benchmark datasets while reducing time complexity and the number of trainable parameters by 70.6% on average. Compared to another maximally powerful method named Graph Isomporhic Network, we show that we outperform them on two popular benchmark datasets while reducing the number of learnable parameters on average by 60.9%.

15 pages, presented at 21st International Workshop on Mining and Learning with Graphs, AstraZenica Bio & Healthcare award Paper, ECML PKDD 2024 Vilnius

We present an innovative framework for traffic dynamics analysis using High-Order Evolving Graphs, designed to improve spatio-temporal representations in autonomous driving contexts. Our approach constructs temporal bidirectional bipartite graphs that effectively model the complex interactions within traffic scenes in real-time. By integrating Graph Neural Networks (GNNs) with high-order multi-aggregation strategies, we significantly enhance the modeling of traffic scene dynamics, providing a more accurate and detailed analysis of these interactions. Additionally, we incorporate inductive learning techniques inspired by the GraphSAGE framework, enabling our model to adapt to new and unseen traffic scenarios without the need for retraining, thus ensuring robust generalization. Through extensive experiments on the ROAD and ROAD Waymo datasets, we establish a comprehensive baseline for further developments, demonstrating the potential of our method in accurately capturing traffic behavior. Our results emphasize the value of high-order statistical moments and feature-gated attention mechanisms in improving traffic behavior analysis, laying the groundwork for advancing autonomous driving technologies. Our source code is available at: https://github.com/Addy-1998/High_Order_Graphs

Accepted manuscript - 2nd Workshop on Vision-Centric Autonomous Driving (VCAD) as part of European Conference on Computer Vision (ECCV) 2024

Molecular dynamics simulations offer detailed insights into atomic motions but face timescale limitations. Enhanced sampling methods have addressed these challenges but even with machine learning, they often rely on pre-selected expert-based features. In this work, we present the Graph Neural Network-State Predictive Information Bottleneck (GNN-SPIB) framework, which combines graph neural networks and the State Predictive Information Bottleneck to automatically learn low-dimensional representations directly from atomic coordinates. Tested on three benchmark systems, our approach predicts essential structural, thermodynamic and kinetic information for slow processes, demonstrating robustness across diverse systems. The method shows promise for complex systems, enabling effective enhanced sampling without requiring pre-defined reaction coordinates or input features.

Temporal graph neural networks (TGNNs) outperform regular GNNs by incorporating time information into graph-based operations. However, TGNNs adopt specialized models (e.g., TGN, TGAT, and APAN ) and require tailored training frameworks (e.g., TGL and ETC). In this paper, we propose TF-TGN, which uses Transformer decoder as the backbone model for TGNN to enjoy Transformer's codebase for efficient training. In particular, Transformer achieves tremendous success for language modeling, and thus the community developed high-performance kernels (e.g., flash-attention and memory-efficient attention) and efficient distributed training schemes (e.g., PyTorch FSDP, DeepSpeed, and Megatron-LM). We observe that TGNN resembles language modeling, i.e., the message aggregation operation between chronologically occurring nodes and their temporal neighbors in TGNNs can be structured as sequence modeling. Beside this similarity, we also incorporate a series of algorithm designs including suffix infilling, temporal graph attention with self-loop, and causal masking self-attention to make TF-TGN work. During training, existing systems are slow in transforming the graph topology and conducting graph sampling. As such, we propose methods to parallelize the CSR format conversion and graph sampling. We also adapt Transformer codebase to train TF-TGN efficiently with multiple GPUs. We experiment with 9 graphs and compare with 2 state-of-the-art TGNN training frameworks. The results show that TF-TGN can accelerate training by over 2.20 while providing comparable or even superior accuracy to existing SOTA TGNNs. TF-TGN is available at https://github.com/qianghuangwhu/TF-TGN.

conference Under review

Graph Neural Networks (GNNs) have achieved notable success in the analysis of non-Euclidean data across a wide range of domains. However, their applicability is constrained by the dependence on the observed graph structure. To solve this problem, Latent Graph Inference (LGI) is proposed to infer a task-specific latent structure by computing similarity or edge probability of node features and then apply a GNN to produce predictions. Even so, existing approaches neglect the noise from node features, which affects generated graph structure and performance. In this work, we introduce a novel method called Probability Passing to refine the generated graph structure by aggregating edge probabilities of neighboring nodes based on observed graph. Furthermore, we continue to utilize the LGI framework, inputting the refined graph structure and node features into GNNs to obtain predictions. We name the proposed scheme as Probability Passing-based Graph Neural Network (PPGNN). Moreover, the anchor-based technique is employed to reduce complexity and improve efficiency. Experimental results demonstrate the effectiveness of the proposed method.

Graph machine learning, particularly using graph neural networks, fundamentally relies on node features. Nevertheless, numerous real-world systems, such as social and biological networks, often lack node features due to various reasons, including privacy concerns, incomplete or missing data, and limitations in data collection. In such scenarios, researchers typically resort to methods like structural and positional encoding to construct node features. However, the length of such features is contingent on the maximum value within the property being encoded, for example, the highest node degree, which can be exceedingly large in applications like scale-free networks. Furthermore, these encoding schemes are limited to categorical data and might not be able to encode metrics returning other type of values. In this paper, we introduce a novel, universally applicable encoder, termed PropEnc, which constructs expressive node embedding from any given graph metric. PropEnc leverages histogram construction combined with reverse index encoding, offering a flexible method for node features initialization. It supports flexible encoding in terms of both dimensionality and type of input, demonstrating its effectiveness across diverse applications. PropEnc allows encoding metrics in low-dimensional space which effectively avoids the issue of sparsity and enhances the efficiency of the models. We show that \emph{PropEnc} can construct node features that either exactly replicate one-hot encoding or closely approximate indices under various settings. Our extensive evaluations in graph classification setting across multiple social networks that lack node features support our hypothesis. The empirical results conclusively demonstrate that PropEnc is both an efficient and effective mechanism for constructing node features from diverse set of graph metrics.

Our study reveals new theoretical insights into over-smoothing and feature over-correlation in graph neural networks. Specifically, we demonstrate that with increased depth, node representations become dominated by a low-dimensional subspace that depends on the aggregation function but not on the feature transformations. For all aggregation functions, the rank of the node representations collapses, resulting in over-smoothing for particular aggregation functions. Our study emphasizes the importance for future research to focus on rank collapse rather than over-smoothing. Guided by our theory, we propose a sum of Kronecker products as a beneficial property that provably prevents over-smoothing, over-correlation, and rank collapse. We empirically demonstrate the shortcomings of existing models in fitting target functions of node classification tasks.

A graph neural network (GNN) approach is introduced in this work which enables mesh-based three-dimensional super-resolution of fluid flows. In this framework, the GNN is designed to operate not on the full mesh-based field at once, but on localized meshes of elements (or cells) directly. To facilitate mesh-based GNN representations in a manner similar to spectral (or finite) element discretizations, a baseline GNN layer (termed a message passing layer, which updates local node properties) is modified to account for synchronization of coincident graph nodes, rendering compatibility with commonly used element-based mesh connectivities. The architecture is multiscale in nature, and is comprised of a combination of coarse-scale and fine-scale message passing layer sequences (termed processors) separated by a graph unpooling layer. The coarse-scale processor embeds a query element (alongside a set number of neighboring coarse elements) into a single latent graph representation using coarse-scale synchronized message passing over the element neighborhood, and the fine-scale processor leverages additional message passing operations on this latent graph to correct for interpolation errors. Demonstration studies are performed using hexahedral mesh-based data from Taylor-Green Vortex flow simulations at Reynolds numbers of 1600 and 3200. Through analysis of both global and local errors, the results ultimately show how the GNN is able to produce accurate super-resolved fields compared to targets in both coarse-scale and multiscale model configurations.

When employing deep neural networks (DNNs) for semantic segmentation in safety-critical applications like automotive perception or medical imaging, it is important to estimate their performance at runtime, e.g. via uncertainty estimates or prediction quality estimates. Previous works mostly performed uncertainty estimation on pixel-level. In a line of research, a connected-component-wise (segment-wise) perspective was taken, approaching uncertainty estimation on an object-level by performing so-called meta classification and regression to estimate uncertainty and prediction quality, respectively. In those works, each predicted segment is considered individually to estimate its uncertainty or prediction quality. However, the neighboring segments may provide additional hints on whether a given predicted segment is of high quality, which we study in the present work. On the basis of uncertainty indicating metrics on segment-level, we use graph neural networks (GNNs) to model the relationship of a given segment's quality as a function of the given segment's metrics as well as those of its neighboring segments. We compare different GNN architectures and achieve a notable performance improvement.

11 pages, 3 figures, submitted to BMVC "Workshop on Robust Recognition in the Open World" (https://rrow2024.github.io/call-for-papers)

Proactive content moderation requires platforms to rapidly and continuously evaluate the credibility of websites. Leveraging the direct and indirect paths users follow to unreliable websites, we develop a website credibility classification and discovery system that integrates both webgraph and large-scale social media contexts. We additionally introduce the concept of dredge words, terms or phrases for which unreliable domains rank highly on search engines, and provide the first exploration of their usage on social media. Our graph neural networks that combine webgraph and social media contexts generate to state-of-the-art results in website credibility classification and significantly improves the top-k identification of unreliable domains. Additionally, we release a novel dataset of dredge words, highlighting their strong connections to both social media and online commerce platforms.

We present a Morphology-Informed Heterogeneous Graph Neural Network (MI-HGNN) for learning-based contact perception. The architecture and connectivity of the MI-HGNN are constructed from the robot morphology, in which nodes and edges are robot joints and links, respectively. By incorporating the morphology-informed constraints into a neural network, we improve a learning-based approach using model-based knowledge. We apply the proposed MI-HGNN to two contact perception problems, and conduct extensive experiments using both real-world and simulated data collected using two quadruped robots. Our experiments demonstrate the superiority of our method in terms of effectiveness, generalization ability, model efficiency, and sample efficiency. Our MI-HGNN improved the performance of a state-of-the-art model that leverages robot morphological symmetry by 8.4% with only 0.21% of its parameters. Although MI-HGNN is applied to contact perception problems for legged robots in this work, it can be seamlessly applied to other types of multi-body dynamical systems and has the potential to improve other robot learning frameworks. Our code is made publicly available at https://github.com/lunarlab-gatech/Morphology-Informed-HGNN.

6 pages, 5 figures; This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

Graph neural networks (GNNs) are a type of neural network capable of learning on graph-structured data. However, training GNNs on large-scale graphs is challenging due to iterative aggregations of high-dimensional features from neighboring vertices within sparse graph structures combined with neural network operations. The sparsity of graphs frequently results in suboptimal memory access patterns and longer training time. Graph reordering is an optimization strategy aiming to improve the graph data layout. It has shown to be effective to speed up graph analytics workloads, but its effect on the performance of GNN training has not been investigated yet. The generalization of reordering to GNN performance is nontrivial, as multiple aspects must be considered: GNN hyper-parameters such as the number of layers, the number of hidden dimensions, and the feature size used in the GNN model, neural network operations, large intermediate vertex states, and GPU acceleration. In our work, we close this gap by performing an empirical evaluation of 12 reordering strategies in two state-of-the-art GNN systems, PyTorch Geometric and Deep Graph Library. Our results show that graph reordering is effective in reducing training time for CPU- and GPU-based training, respectively. Further, we find that GNN hyper-parameters influence the effectiveness of reordering, that reordering metrics play an important role in selecting a reordering strategy, that lightweight reordering performs better for GPU-based than for CPU-based training, and that invested reordering time can in many cases be amortized.

To be published in proceedings of the 51st International Conference on Very Large Data Bases (VLDB), September 1-5, 2025

Deep neural networks (DNNs) have demonstrated remarkable performance across various domains, yet their application to temporal graph regression tasks faces significant challenges regarding interpretability. This critical issue, rooted in the inherent complexity of both DNNs and underlying spatio-temporal patterns in the graph, calls for innovative solutions. While interpretability concerns in Graph Neural Networks (GNNs) mirror those of DNNs, to the best of our knowledge, no notable work has addressed the interpretability of temporal GNNs using a combination of Information Bottleneck (IB) principles and prototype-based methods. Our research introduces a novel approach that uniquely integrates these techniques to enhance the interpretability of temporal graph regression models. The key contributions of our work are threefold: We introduce the \underline{G}raph \underline{IN}terpretability in \underline{T}emporal \underline{R}egression task using \underline{I}nformation bottleneck and \underline{P}rototype (GINTRIP) framework, the first combined application of IB and prototype-based methods for interpretable temporal graph tasks. We derive a novel theoretical bound on mutual information (MI), extending the applicability of IB principles to graph regression tasks. We incorporate an unsupervised auxiliary classification head, fostering multi-task learning and diverse concept representation, which enhances the model bottleneck's interpretability. Our model is evaluated on real-world traffic datasets, outperforming existing methods in both forecasting accuracy and interpretability-related metrics.

This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible

Understanding neurological disorder is a fundamental problem in neuroscience, which often requires the analysis of brain networks derived from functional magnetic resonance imaging (fMRI) data. Despite the prevalence of Graph Neural Networks (GNNs) and Graph Transformers in various domains, applying them to brain networks faces challenges. Specifically, the datasets are severely impacted by the noises caused by distribution shifts across sub-populations and the neglect of node identities, both obstruct the identification of disease-specific patterns. To tackle these challenges, we propose Contrasformer, a novel contrastive brain network Transformer. It generates a prior-knowledge-enhanced contrast graph to address the distribution shifts across sub-populations by a two-stream attention mechanism. A cross attention with identity embedding highlights the identity of nodes, and three auxiliary losses ensure group consistency. Evaluated on 4 functional brain network datasets over 4 different diseases, Contrasformer outperforms the state-of-the-art methods for brain networks by achieving up to 10.8\% improvement in accuracy, which demonstrates its efficacy in neurological disorder identification. Case studies illustrate its interpretability, especially in the context of neuroscience. This paper provides a solution for analyzing brain networks, offering valuable insights into neurological disorders. Our code is available at \url{https://github.com/AngusMonroe/Contrasformer}.

While Graph Neural Networks (GNNs) have achieved enormous success in multiple graph analytical tasks, modern variants mostly rely on the strong inductive bias of homophily. However, real-world networks typically exhibit both homophilic and heterophilic linking patterns, wherein adjacent nodes may share dissimilar attributes and distinct labels. Therefore, GNNs smoothing node proximity holistically may aggregate both task-relevant and irrelevant (even harmful) information, limiting their ability to generalize to heterophilic graphs and potentially causing non-robustness. In this work, we propose a novel Edge Splitting GNN (ES-GNN) framework to adaptively distinguish between graph edges either relevant or irrelevant to learning tasks. This essentially transfers the original graph into two subgraphs with the same node set but complementary edge sets dynamically. Given that, information propagation separately on these subgraphs and edge splitting are alternatively conducted, thus disentangling the task-relevant and irrelevant features. Theoretically, we show that our ES-GNN can be regarded as a solution to a disentangled graph denoising problem, which further illustrates our motivations and interprets the improved generalization beyond homophily. Extensive experiments over 11 benchmark and 1 synthetic datasets not only demonstrate the effective performance of ES-GNN but also highlight its robustness to adversarial graphs and mitigation of the over-smoothing problem.

Self-exciting point processes are widely used to model the contagious effects of crime events living within continuous geographic space, using their occurrence time and locations. However, in urban environments, most events are naturally constrained within the city's street network structure, and the contagious effects of crime are governed by such a network geography. Meanwhile, the complex distribution of urban infrastructures also plays an important role in shaping crime patterns across space. We introduce a novel spatio-temporal-network point process framework for crime modeling that integrates these urban environmental characteristics by incorporating self-attention graph neural networks. Our framework incorporates the street network structure as the underlying event space, where crime events can occur at random locations on the network edges. To realistically capture criminal movement patterns, distances between events are measured using street network distances. We then propose a new mark for a crime event by concatenating the event's crime category with the type of its nearby landmark, aiming to capture how the urban design influences the mixing structures of various crime types. A graph attention network architecture is adopted to learn the existence of mark-to-mark interactions. Extensive experiments on crime data from Valencia, Spain, demonstrate the effectiveness of our framework in understanding the crime landscape and forecasting crime risks across regions.

The integration of power electronics building blocks in modern MVDC 12kV Naval ship systems enhances energy management and functionality but also introduces complex fault detection and control challenges. These challenges strain traditional fault diagnostic methods, making it difficult to detect and manage faults across multiple locations while maintaining system stability and performance. This paper proposes a temporal recurrent graph transformer network for fault diagnosis in naval MVDC 12kV shipboard systems. The deep graph neural network uses gated recurrent units to capture temporal features and a multi-head attention mechanism to extract spatial features, enhancing diagnostic accuracy. The approach effectively identifies and evaluates successive multiple faults with high precision. The method is implemented and validated on the MVDC 12kV shipboard system designed by the ESDRC team, incorporating all key components. Results show significant improvements in fault localization accuracy, with a 1-4% increase in performance metrics compared to other machine learning methods.

Connected multi-agent robotic systems (MRS) are prone to deadlocks in an obstacle environment where the robots can get stuck away from their desired locations under a smooth low-level control policy. Without an external intervention, often in terms of a high-level command, a low-level control policy cannot resolve such deadlocks. Utilizing the generalizability and low data requirements of foundation models, this paper explores the possibility of using text-based models, i.e., large language models (LLMs), and text-and-image-based models, i.e., vision-language models (VLMs), as high-level planners for deadlock resolution. We propose a hierarchical control framework where a foundation model-based high-level planner helps to resolve deadlocks by assigning a leader to the MRS along with a set of waypoints for the MRS leader. Then, a low-level distributed control policy based on graph neural networks is executed to safely follow these waypoints, thereby evading the deadlock. We conduct extensive experiments on various MRS environments using the best available pre-trained LLMs and VLMs. We compare their performance with a graph-based planner in terms of effectiveness in helping the MRS reach their target locations and computational time. Our results illustrate that, compared to grid-based planners, the foundation models perform better in terms of the goal-reaching rate and computational time for complex environments, which helps us conclude that foundation models can assist MRS operating in complex obstacle-cluttered environments to resolve deadlocks efficiently.

Autoencoders based on Graph Neural Networks (GNNs) have garnered significant attention in recent years for their ability to extract informative latent representations, characterizing the structure of complex topologies, such as graphs. Despite the prevalence of Graph Autoencoders, there has been limited focus on developing and evaluating explainable neural-based graph generative models specifically designed for signed networks. To address this gap, we propose the Signed Graph Archetypal Autoencoder (SGAAE) framework. SGAAE extracts node-level representations that express node memberships over distinct extreme profiles, referred to as archetypes, within the network. This is achieved by projecting the graph onto a learned polytope, which governs its polarization. The framework employs a recently proposed likelihood for analyzing signed networks based on the Skellam distribution, combined with relational archetypal analysis and GNNs. Our experimental evaluation demonstrates the SGAAEs' capability to successfully infer node memberships over the different underlying latent structures while extracting competing communities formed through the participation of the opposing views in the network. Additionally, we introduce the 2-level network polarization problem and show how SGAAE is able to characterize such a setting. The proposed model achieves high performance in different tasks of signed link prediction across four real-world datasets, outperforming several baseline models.

Top-k algorithms are essential in various applications, from high-performance computing and information retrieval to big data and neural network model training. This paper introduces RTop-K, a highly efficient parallel row-wise top-k selection algorithm designed for GPUs. RTop-K employs a Binary Search-based approach to optimize resource allocation and provides a scalable solution that significantly accelerates top-k operations. We perform a theoretical analysis of the effects of early stopping in our algorithm, demonstrating that it maintains the accuracy of neural network models while enhancing performance. Comprehensive tests show that our GPU implementation of RTop-K outperforms other row-wise top-k GPU implementations, with minimal impact on testing accuracy when early stopping is applied. Notably, RTop-K achieves speed increases ranging from 4.245$\times$ to 9.506$\times$ with early stopping, and 3.936$\times$ without early stopping, compared to state-of-the-art implementations. The proposed methods offer significant improvements in the training and inference of Graph Neural Networks (GNNs), addressing critical challenges in latency and throughput on GPU platforms.

Need to improve the experiment part

In the domain of point cloud analysis, despite the significant capabilities of Graph Neural Networks (GNNs) in managing complex 3D datasets, existing approaches encounter challenges like high computational costs and scalability issues with extensive scenarios. These limitations restrict the practical deployment of GNNs, notably in resource-constrained environments. To address these issues, this study introduce Point<\b> Vi<\b>sion G<\b>NN (PointViG), an efficient framework for point cloud analysis. PointViG incorporates a lightweight graph convolutional module to efficiently aggregate local features and mitigate over-smoothing. For large-scale point cloud scenes, we propose an adaptive dilated graph convolution technique that searches for sparse neighboring nodes within a dilated neighborhood based on semantic correlation, thereby expanding the receptive field and ensuring computational efficiency. Experiments demonstrate that PointViG achieves performance comparable to state-of-the-art models while balancing performance and complexity. On the ModelNet40 classification task, PointViG achieved 94.3% accuracy with 1.5M parameters. For the S3DIS segmentation task, it achieved an mIoU of 71.7% with 5.3M parameters. These results underscore the potential and efficiency of PointViG in point cloud analysis.

Hypergraphs tackle the limitations of traditional graphs by introducing {\em hyperedges}. While graph edges connect only two nodes, hyperedges connect an arbitrary number of nodes along their edges. Also, the underlying message-passing mechanisms in Hypergraph Neural Networks (HGNNs) are in the form of vertex-hyperedge-vertex, which let HGNNs capture and utilize richer and more complex structural information than traditional Graph Neural Networks (GNNs). More recently, the idea of overlapping subgraphs has emerged. These subgraphs can capture more information about subgroups of vertices without limiting one vertex belonging to just one group, allowing vertices to belong to multiple groups or subgraphs. In addition, one of the most important problems in graph clustering is to find densest overlapping subgraphs (DOS). In this paper, we propose a solution to the DOS problem via Agglomerative Greedy Enumeration (DOSAGE) algorithm as a novel approach to enhance the process of generating the densest overlapping subgraphs and, hence, a robust construction of the hypergraphs. Experiments on standard benchmarks show that the DOSAGE algorithm significantly outperforms the HGNNs and six other methods on the node classification task.

Whilst spectral Graph Neural Networks (GNNs) are theoretically well-founded in the spectral domain, their practical reliance on polynomial approximation implies a profound linkage to the spatial domain. As previous studies rarely examine spectral GNNs from the spatial perspective, their spatial-domain interpretability remains elusive, e.g., what information is essentially encoded by spectral GNNs in the spatial domain? In this paper, to answer this question, we establish a theoretical connection between spectral filtering and spatial aggregation, unveiling an intrinsic interaction that spectral filtering implicitly leads the original graph to an adapted new graph, explicitly computed for spatial aggregation. Both theoretical and empirical investigations reveal that the adapted new graph not only exhibits non-locality but also accommodates signed edge weights to reflect label consistency among nodes. These findings thus highlight the interpretable role of spectral GNNs in the spatial domain and inspire us to rethink graph spectral filters beyond the fixed-order polynomials, which neglect global information. Built upon the theoretical findings, we revisit the state-of-the-art spectral GNNs and propose a novel Spatially Adaptive Filtering (SAF) framework, which leverages the adapted new graph by spectral filtering for an auxiliary non-local aggregation. Notably, our proposed SAF comprehensively models both node similarity and dissimilarity from a global perspective, therefore alleviating persistent deficiencies of GNNs related to long-range dependencies and graph heterophily. Extensive experiments over 13 node classification benchmarks demonstrate the superiority of our proposed framework to the state-of-the-art models.

Despite the promising performance of state of the art approaches for Parkinsons Disease (PD) detection, these approaches often analyze individual speech segments in isolation, which can lead to suboptimal results. Dysarthric cues that characterize speech impairments from PD patients are expected to be related across segments from different speakers. Isolated segment analysis fails to exploit these inter segment relationships. Additionally, not all speech segments from PD patients exhibit clear dysarthric symptoms, introducing label noise that can negatively affect the performance and generalizability of current approaches. To address these challenges, we propose a novel PD detection framework utilizing Graph Convolutional Networks (GCNs). By representing speech segments as nodes and capturing the similarity between segments through edges, our GCN model facilitates the aggregation of dysarthric cues across the graph, effectively exploiting segment relationships and mitigating the impact of label noise. Experimental results demonstrate theadvantages of the proposed GCN model for PD detection and provide insights into its underlying mechanisms

Submitted to ICASSP 2025

The rise of online social networks has facilitated the evolution of social recommender systems, which incorporate social relations to enhance users' decision-making process. With the great success of Graph Neural Networks (GNNs) in learning node representations, GNN-based social recommendations have been widely studied to model user-item interactions and user-user social relations simultaneously. Despite their great successes, recent studies have shown that these advanced recommender systems are highly vulnerable to adversarial attacks, in which attackers can inject well-designed fake user profiles to disrupt recommendation performances. While most existing studies mainly focus on argeted attacks to promote target items on vanilla recommender systems, untargeted attacks to degrade the overall prediction performance are less explored on social recommendations under a black-box scenario. To perform untargeted attacks on social recommender systems, attackers can construct malicious social relationships for fake users to enhance the attack performance. However, the coordination of social relations and item profiles is challenging for attacking black-box social recommendations. To address this limitation, we first conduct several preliminary studies to demonstrate the effectiveness of cross-community connections and cold-start items in degrading recommendations performance. Specifically, we propose a novel framework MultiAttack based on multi-agent reinforcement learning to coordinate the generation of cold-start item profiles and cross-community social relations for conducting untargeted attacks on black-box social recommendations. Comprehensive experiments on various real-world datasets demonstrate the effectiveness of our proposed attacking framework under the black-box setting.

Graph anomaly detection (GAD), which aims to identify unusual graph instances (nodes, edges, subgraphs, or graphs), has attracted increasing attention in recent years due to its significance in a wide range of applications. Deep learning approaches, graph neural networks (GNNs) in particular, have been emerging as a promising paradigm for GAD, owing to its strong capability in capturing complex structure and/or node attributes in graph data. Considering the large number of methods proposed for GNN-based GAD, it is of paramount importance to summarize the methodologies and findings in the existing GAD studies, so that we can pinpoint effective model designs for tackling open GAD problems. To this end, in this work we aim to present a comprehensive review of deep learning approaches for GAD. Existing GAD surveys are focused on task-specific discussions, making it difficult to understand the technical insights of existing methods and their limitations in addressing some unique challenges in GAD. To fill this gap, we first discuss the problem complexities and their resulting challenges in GAD, and then provide a systematic review of current deep GAD methods from three novel perspectives of methodology, including GNN backbone design, proxy task design for GAD, and graph anomaly measures. To deepen the discussions, we further propose a taxonomy of 13 fine-grained method categories under these three perspectives to provide more in-depth insights into the model designs and their capabilities. To facilitate the experiments and validation, we also summarize a collection of widely-used GAD datasets and empirical comparison. We further discuss multiple open problems to inspire more future high-quality research. A continuously updated repository for datasets, links to the codes of algorithms, and empirical comparison is available at https://github.com/mala-lab/Awesome-Deep-Graph-Anomaly-Detection.

24 pages, 6 figures, and 7 tables

This paper introduces SGCode, a flexible prompt-optimizing system to generate secure code with large language models (LLMs). SGCode integrates recent prompt-optimization approaches with LLMs in a unified system accessible through front-end and back-end APIs, enabling users to 1) generate secure code, which is free of vulnerabilities, 2) review and share security analysis, and 3) easily switch from one prompt optimization approach to another, while providing insights on model and system performance. We populated SGCode on an AWS server with PromSec, an approach that optimizes prompts by combining an LLM and security tools with a lightweight generative adversarial graph neural network to detect and fix security vulnerabilities in the generated code. Extensive experiments show that SGCode is practical as a public tool to gain insights into the trade-offs between model utility, secure code generation, and system cost. SGCode has only a marginal cost compared with prompting LLMs. SGCode is available at: http://3.131.141.63:8501/.

Machine-learned force fields (MLFFs) promise to offer a computationally efficient alternative to ab initio simulations for complex molecular systems. However, ensuring their generalizability beyond training data is crucial for their wide application in studying solid materials. This work investigates the ability of a graph neural network (GNN)-based MLFF, trained on Lennard-Jones Argon, to describe solid-state phenomena not explicitly included during training. We assess the MLFF's performance in predicting phonon density of states (PDOS) for a perfect face-centered cubic (FCC) crystal structure at both zero and finite temperatures. Additionally, we evaluate vacancy migration rates and energy barriers in an imperfect crystal using direct molecular dynamics (MD) simulations and the string method. Notably, vacancy configurations were absent from the training data. Our results demonstrate the MLFF's capability to capture essential solid-state properties with good agreement to reference data, even for unseen configurations. We further discuss data engineering strategies to enhance the generalizability of MLFFs. The proposed set of benchmark tests and workflow for evaluating MLFF performance in describing perfect and imperfect crystals pave the way for reliable application of MLFFs in studying complex solid-state materials.

Financial fraud refers to the act of obtaining financial benefits through dishonest means. Such behavior not only disrupts the order of the financial market but also harms economic and social development and breeds other illegal and criminal activities. With the popularization of the internet and online payment methods, many fraudulent activities and money laundering behaviors in life have shifted from offline to online, posing a great challenge to regulatory authorities. How to efficiently detect these financial fraud activities has become an urgent issue that needs to be resolved. Graph neural networks are a type of deep learning model that can utilize the interactive relationships within graph structures, and they have been widely applied in the field of fraud detection. However, there are still some issues. First, fraudulent activities only account for a very small part of transaction transfers, leading to an inevitable problem of label imbalance in fraud detection. At the same time, fraudsters often disguise their behavior, which can have a negative impact on the final prediction results. In addition, existing research has overlooked the importance of balancing neighbor information and central node information. For example, when the central node has too many neighbors, the features of the central node itself are often neglected. Finally, fraud activities and patterns are constantly changing over time, so considering the dynamic evolution of graph edge relationships is also very important.

The ability of Graph Neural Networks (GNNs) to capture long-range and global topology information is limited by the scope of conventional graph Laplacian, leading to unsatisfactory performance on some datasets, particularly on heterophilic graphs. To address this limitation, we propose a new class of parameterized Laplacian matrices, which provably offers more flexibility in controlling the diffusion distance between nodes than the conventional graph Laplacian, allowing long-range information to be adaptively captured through diffusion on graph. Specifically, we first prove that the diffusion distance and spectral distance on graph have an order-preserving relationship. With this result, we demonstrate that the parameterized Laplacian can accelerate the diffusion of long-range information, and the parameters in the Laplacian enable flexibility of the diffusion scopes. Based on the theoretical results, we propose topology-guided rewiring mechanism to capture helpful long-range neighborhood information for heterophilic graphs. With this mechanism and the new Laplacian, we propose two GNNs with flexible diffusion scopes: namely the Parameterized Diffusion based Graph Convolutional Networks (PD-GCN) and Graph Attention Networks (PD-GAT). Synthetic experiments reveal the high correlations between the parameters of the new Laplacian and the performance of parameterized GNNs under various graph homophily levels, which verifies that our new proposed GNNs indeed have the ability to adjust the parameters to adaptively capture the global information for different levels of heterophilic graphs. They also outperform the state-of-the-art (SOTA) models on 6 out of 7 real-world benchmark datasets, which further confirms their superiority.

arXiv admin note: substantial text overlap with arXiv:2403.01475

Building-specific knowledge such as building type and function information is important for numerous energy applications. However, comprehensive datasets containing this information for individual households are missing in many regions of Europe. For the first time, we investigate whether it is feasible to predict building types and functional classes at a European scale based on only open GIS datasets available across countries. We train a graph neural network (GNN) classifier on a large-scale graph dataset consisting of OpenStreetMap (OSM) buildings across the EU, Norway, Switzerland, and the UK. To efficiently perform training using the large-scale graph, we utilize localized subgraphs. A graph transformer model achieves a high Cohen's kappa coefficient of 0.754 when classifying buildings into 9 classes, and a very high Cohen's kappa coefficient of 0.844 when classifying buildings into the residential and non-residential classes. The experimental results imply three core novel contributions to literature. Firstly, we show that building classification across multiple countries is possible using a multi-source dataset consisting of information about 2D building shape, land use, degree of urbanization, and countries as input, and OSM tags as ground truth. Secondly, our results indicate that GNN models that consider contextual information about building neighborhoods improve predictive performance compared to models that only consider individual buildings and ignore the neighborhood. Thirdly, we show that training with GNNs on localized subgraphs instead of standard GNNs improves performance for the task of building classification.

Power flow analysis plays a critical role in the control and operation of power systems. The high computational burden of traditional solution methods led to a shift towards data-driven approaches, exploiting the availability of digital metering data. However, data-driven approaches, such as deep learning, have not yet won the trust of operators as they are agnostic to the underlying physical model and have poor performances in regimes with limited observability. To address these challenges, this paper proposes a new, physics-informed model. More specifically, a novel physics-informed loss function is developed that can be used to train (deep) neural networks aimed at power flow simulation. The loss function is not only based on the theoretical AC power flow equations that govern the problem but also incorporates real physical line losses, resulting in higher loss accuracy and increased learning potential. The proposed model is used to train a Graph Neural Network (GNN) and is evaluated on a small 3-bus test case both against another physics-informed GNN that does not incorporate physical losses and against a model-free technique. The validation results show that the proposed model outperforms the conventional physics-informed network on all used performance metrics. Even more interesting is that the model shows strong prediction capabilities when tested on scenarios outside the training sample set, something that is a substantial deficiency of model-free techniques.

Quantifying uncertainty is crucial for robust and reliable predictions. However, existing spatiotemporal deep learning mostly focuses on deterministic prediction, overlooking the inherent uncertainty in such prediction. Particularly, highly-granular spatiotemporal datasets are often sparse, posing extra challenges in prediction and uncertainty quantification. To address these issues, this paper introduces a novel post-hoc Sparsity-awar Uncertainty Calibration (SAUC) framework, which calibrates uncertainty in both zero and non-zero values. To develop SAUC, we firstly modify the state-of-the-art deterministic spatiotemporal Graph Neural Networks (ST-GNNs) to probabilistic ones in the pre-calibration phase. Then we calibrate the probabilistic ST-GNNs for zero and non-zero values using quantile approaches.Through extensive experiments, we demonstrate that SAUC can effectively fit the variance of sparse data and generalize across two real-world spatiotemporal datasets at various granularities. Specifically, our empirical experiments show a 20\% reduction in calibration errors in zero entries on the sparse traffic accident and urban crime prediction. Overall, this work demonstrates the theoretical and empirical values of the SAUC framework, thus bridging a significant gap between uncertainty quantification and spatiotemporal prediction.

Paper accepted by ACM SIGSPATIAL 2024

Graph neural networks (GNNs) have become a workhorse approach for learning from data defined over irregular domains, typically by implicitly assuming that the data structure is represented by a homophilic graph. However, recent works have revealed that many relevant applications involve heterophilic data where the performance of GNNs can be notably compromised. To address this challenge, we present a simple yet effective architecture designed to mitigate the limitations of the homophily assumption. The proposed architecture reinterprets the role of graph filters in convolutional GNNs, resulting in a more general architecture while incorporating a stronger inductive bias than GNNs based on filter banks. The proposed convolutional layer enhances the expressive capacity of the architecture enabling it to learn from both homophilic and heterophilic data and preventing the issue of oversmoothing. From a theoretical standpoint, we show that the proposed architecture is permutation equivariant. Finally, we show that the proposed GNNs compares favorably relative to several state-of-the-art baselines in both homophilic and heterophilic datasets, showcasing its promising potential.

This paper presents SYBILGAT, a novel approach to Sybil detection in social networks using Graph Attention Networks (GATs). Traditional methods for Sybil detection primarily leverage structural properties of networks; however, they tend to struggle with a large number of attack edges and are often unable to simultaneously utilize both known Sybil and honest nodes. Our proposed method addresses these limitations by dynamically assigning attention weights to different nodes during aggregations, enhancing detection performance. We conducted extensive experiments in various scenarios, including pretraining in sampled subgraphs, synthetic networks, and networks under targeted attacks. The results show that SYBILGAT significantly outperforms the state-of-the-art algorithms, particularly in scenarios with high attack complexity and when the number of attack edges increases. Our approach shows robust performance across different network models and sizes, even as the detection task becomes more challenging. We successfully applied the model to a real-world Twitter graph with more than 269k nodes and 6.8M edges. The flexibility and generalizability of SYBILGAT make it a promising tool to defend against Sybil attacks in online social networks with only structural information.

9 pages, 1 figure, 6 tables

Graph Neural Networks (GNNs) have been widely employed for feature representation learning in molecular graphs. Therefore, it is crucial to enhance the expressiveness of feature representation to ensure the effectiveness of GNNs. However, a significant portion of current research primarily focuses on the structural features within individual molecules, often overlooking the structural similarity between molecules, which is a crucial aspect encapsulating rich information on the relationship between molecular properties and structural characteristics. Thus, these approaches fail to capture the rich semantic information at the molecular structure level. To bridge this gap, we introduce the \textbf{Molecular Structural Similarity Motif GNN (MSSM-GNN)}, a novel molecular graph representation learning method that can capture structural similarity information among molecules from a global perspective. In particular, we propose a specially designed graph that leverages graph kernel algorithms to represent the similarity between molecules quantitatively. Subsequently, we employ GNNs to learn feature representations from molecular graphs, aiming to enhance the accuracy of property prediction by incorporating additional molecular representation information. Finally, through a series of experiments conducted on both small-scale and large-scale molecular datasets, we demonstrate that our model consistently outperforms eleven state-of-the-art baselines. The codes are available at https://github.com/yaoyao-yaoyao-cell/MSSM-GNN.

As network data applications continue to expand, causal inference within networks has garnered increasing attention. However, hidden confounders complicate the estimation of causal effects. Most methods rely on the strong ignorability assumption, which presumes the absence of hidden confounders-an assumption that is both difficult to validate and often unrealistic in practice. To address this issue, we propose CgNN, a novel approach that leverages network structure as instrumental variables (IVs), combined with graph neural networks (GNNs) and attention mechanisms, to mitigate hidden confounder bias and improve causal effect estimation. By utilizing network structure as IVs, we reduce confounder bias while preserving the correlation with treatment. Our integration of attention mechanisms enhances robustness and improves the identification of important nodes. Validated on two real-world datasets, our results demonstrate that CgNN effectively mitigates hidden confounder bias and offers a robust GNN-driven IV framework for causal inference in complex network data.

Recommender Systems (RS) play a pivotal role in boosting user satisfaction by providing personalized product suggestions in domains such as e-commerce and entertainment. This study examines the integration of multimodal data text and audio into large language models (LLMs) with the aim of enhancing recommendation performance. Traditional text and audio recommenders encounter limitations such as the cold-start problem, and recent advancements in LLMs, while promising, are computationally expensive. To address these issues, Low-Rank Adaptation (LoRA) is introduced, which enhances efficiency without compromising performance. The ATFLRec framework is proposed to integrate audio and text modalities into a multimodal recommendation system, utilizing various LoRA configurations and modality fusion techniques. Results indicate that ATFLRec outperforms baseline models, including traditional and graph neural network-based approaches, achieving higher AUC scores. Furthermore, separate fine-tuning of audio and text data with distinct LoRA modules yields optimal performance, with different pooling methods and Mel filter bank numbers significantly impacting performance. This research offers valuable insights into optimizing multimodal recommender systems and advancing the integration of diverse data modalities in LLMs.

Estimating causal effects is crucial for decision-makers in many applications, but it is particularly challenging with observational network data due to peer interactions. Many algorithms have been proposed to estimate causal effects involving network data, particularly peer effects, but they often overlook the variety of peer effects. To address this issue, we propose a general setting which considers both peer direct effects and peer indirect effects, and the effect of an individual's own treatment, and provide identification conditions of these causal effects and proofs. To estimate these causal effects, we utilize attention mechanisms to distinguish the influences of different neighbors and explore high-order neighbor effects through multi-layer graph neural networks (GNNs). Additionally, to control the dependency between node features and representations, we incorporate the Hilbert-Schmidt Independence Criterion (HSIC) into the GNN, fully utilizing the structural information of the graph, to enhance the robustness and accuracy of the model. Extensive experiments on two semi-synthetic datasets confirm the effectiveness of our approach. Our theoretical findings have the potential to improve intervention strategies in networked systems, with applications in areas such as social networks and epidemiology.

In the rapidly evolving domain of satellite communications, integrating advanced machine learning techniques, particularly split learning, is crucial for enhancing data processing and model training efficiency across satellites, space stations, and ground stations. Traditional ML approaches often face significant challenges within satellite networks due to constraints such as limited bandwidth and computational resources. To address this gap, we propose a novel framework for more efficient SL in satellite communications. Our approach, Dynamic Topology Informed Pruning, namely DTIP, combines differential privacy with graph and model pruning to optimize graph neural networks for distributed learning. DTIP strategically applies differential privacy to raw graph data and prunes GNNs, thereby optimizing both model size and communication load across network tiers. Extensive experiments across diverse datasets demonstrate DTIP's efficacy in enhancing privacy, accuracy, and computational efficiency. Specifically, on Amazon2M dataset, DTIP maintains an accuracy of 0.82 while achieving a 50% reduction in floating-point operations per second. Similarly, on ArXiv dataset, DTIP achieves an accuracy of 0.85 under comparable conditions. Our framework not only significantly improves the operational efficiency of satellite communications but also establishes a new benchmark in privacy-aware distributed learning, potentially revolutionizing data handling in space-based networks.

To reduce the latency of Backpressure (BP) routing in wireless multi-hop networks, we propose to enhance the existing shortest path-biased BP (SP-BP) and sojourn time-based backlog metrics, since they introduce no additional time step-wise signaling overhead to the basic BP. Rather than relying on hop-distance, we introduce a new edge-weighted shortest path bias built on the scheduling duty cycle of wireless links, which can be predicted by a graph convolutional neural network based on the topology and traffic of wireless networks. Additionally, we tackle three long-standing challenges associated with SP-BP: optimal bias scaling, efficient bias maintenance, and integration of delay awareness. Our proposed solutions inherit the throughput optimality of the basic BP, as well as its practical advantages of low complexity and fully distributed implementation. Our approaches rely on common link features and introduces only a one-time constant overhead to previous SP-BP schemes, or a one-time overhead linear in the network size to the basic BP. Numerical experiments show that our solutions can effectively address the major drawbacks of slow startup, random walk, and the last packet problem in basic BP, improving the end-to-end delay of existing low-overhead BP algorithms under various settings of network traffic, interference, and mobility.

16 pages, 15 figures, accepted for publication in IEEE Transactions on Machine Learning in Communications and Networking. arXiv admin note: text overlap with arXiv:2310.04364, arXiv:2211.10748

In this paper, we develop a data-driven approach to generate incomplete LU factorizations of large-scale sparse matrices. The learned approximate factorization is utilized as a preconditioner for the corresponding linear equation system in the GMRES method. Incomplete factorization methods are one of the most commonly applied algebraic preconditioners for sparse linear equation systems and are able to speed up the convergence of Krylov subspace methods. However, they are sensitive to hyper-parameters and might suffer from numerical breakdown or lead to slow convergence when not properly applied. We replace the typically hand-engineered algorithms with a graph neural network based approach that is trained against data to predict an approximate factorization. This allows us to learn preconditioners tailored for a specific problem distribution. We analyze and empirically evaluate different loss functions to train the learned preconditioners and show their effectiveness to decrease the number of GMRES iterations and improve the spectral properties on our synthetic dataset. The code is available at https://github.com/paulhausner/neural-incomplete-factorization.

The first two authors contributed equally, under review. 12 pages, 5 figures

Graph neural networks have become the default choice by practitioners for graph learning tasks such as graph classification and node classification. Nevertheless, popular graph neural network models still struggle to capture higher-order information, i.e., information that goes \emph{beyond} pairwise interactions. Recent work has shown that persistent homology, a tool from topological data analysis, can enrich graph neural networks with topological information that they otherwise could not capture. Calculating such features is efficient for dimension 0 (connected components) and dimension 1 (cycles). However, when it comes to higher-order structures, it does not scale well, with a complexity of $O(n^d)$, where $n$ is the number of nodes and $d$ is the order of the structures. In this work, we introduce a novel method that extracts information about higher-order structures in the graph while still using the efficient low-dimensional persistent homology algorithm. On standard benchmark datasets, we show that our method can lead to up to $31\%$ improvements in test accuracy.

We introduce Gaussian Garments, a novel approach for reconstructing realistic simulation-ready garment assets from multi-view videos. Our method represents garments with a combination of a 3D mesh and a Gaussian texture that encodes both the color and high-frequency surface details. This representation enables accurate registration of garment geometries to multi-view videos and helps disentangle albedo textures from lighting effects. Furthermore, we demonstrate how a pre-trained graph neural network (GNN) can be fine-tuned to replicate the real behavior of each garment. The reconstructed Gaussian Garments can be automatically combined into multi-garment outfits and animated with the fine-tuned GNN.

Heterogeneous graphs, with nodes and edges of different types, are commonly used to model relational structures in many real-world applications. Standard Graph Neural Networks (GNNs) struggle to process heterogeneous data due to oversmoothing. Instead, current approaches have focused on accounting for the heterogeneity in the model architecture, leading to increasingly complex models. Inspired by recent work, we propose using cellular sheaves to model the heterogeneity in the graph's underlying topology. Instead of modelling the data as a graph, we represent it as cellular sheaves, which allows us to encode the different data types directly in the data structure, eliminating the need to inject them into the architecture. We introduce HetSheaf, a general framework for heterogeneous sheaf neural networks, and a series of heterogeneous sheaf predictors to better encode the data's heterogeneity into the sheaf structure. Finally, we empirically evaluate HetSheaf on several standard heterogeneous graph benchmarks, achieving competitive results whilst being more parameter-efficient.

The problem of multi-task regression over graph nodes has been recently approached through Graph-Instructed Neural Network (GINN), which is a promising architecture belonging to the subset of message-passing graph neural networks. In this work, we discuss the limitations of the Graph-Instructed (GI) layer, and we formalize a novel edge-wise GI (EWGI) layer. We discuss the advantages of the EWGI layer and we provide numerical evidence that EWGINNs perform better than GINNs over graph-structured input data with chaotic connectivity, like the ones inferred from the Erdos-R\'enyi graph.

Terahertz (THz) space communications (Tera-SpaceCom) is envisioned as a promising technology to enable various space science and communication applications. Mainly, the realm of Tera-SpaceCom consists of THz sensing for space exploration, data centers in space providing cloud services for space exploration tasks, and a low earth orbit (LEO) mega-constellation relaying these tasks to ground stations (GSs) or data centers via THz links. Moreover, to reduce the computational burden on data centers as well as resource consumption and latency in the relaying process, the LEO mega-constellation provides satellite edge computing (SEC) services to directly compute space exploration tasks without relaying these tasks to data centers. The LEO satellites that receive space exploration tasks offload (i.e., distribute) partial tasks to their neighboring LEO satellites, to further reduce their computational burden. However, efficient joint communication resource allocation and computing task offloading for the Tera-SpaceCom SEC network is an NP-hard mixed-integer nonlinear programming problem (MINLP), due to the discrete nature of space exploration tasks and sub-arrays as well as the continuous nature of transmit power. To tackle this challenge, a graph neural network (GNN)-deep reinforcement learning (DRL)-based joint resource allocation and task offloading (GRANT) algorithm is proposed with the target of long-term resource efficiency (RE). Particularly, GNNs learn relationships among different satellites from their connectivity information. Furthermore, multi-agent and multi-task mechanisms cooperatively train task offloading and resource allocation. Compared with benchmark solutions, GRANT not only achieves the highest RE with relatively low latency, but realizes the fewest trainable parameters and the shortest running time.

This paper investigates the relationship between graph convolution and Mixup techniques. Graph convolution in a graph neural network involves aggregating features from neighboring samples to learn representative features for a specific node or sample. On the other hand, Mixup is a data augmentation technique that generates new examples by averaging features and one-hot labels from multiple samples. One commonality between these techniques is their utilization of information from multiple samples to derive feature representation. This study aims to explore whether a connection exists between these two approaches. Our investigation reveals that, under two mild conditions, graph convolution can be viewed as a specialized form of Mixup that is applied during both the training and testing phases. The two conditions are: 1) \textit{Homophily Relabel} - assigning the target node's label to all its neighbors, and 2) \textit{Test-Time Mixup} - Mixup the feature during the test time. We establish this equivalence mathematically by demonstrating that graph convolution networks (GCN) and simplified graph convolution (SGC) can be expressed as a form of Mixup. We also empirically verify the equivalence by training an MLP using the two conditions to achieve comparable performance.

Graph representation learning has emerged as a powerful tool for preserving graph topology when mapping nodes to vector representations, enabling various downstream tasks such as node classification and community detection. However, most current graph neural network models face the challenge of requiring extensive labeled data, which limits their practical applicability in real-world scenarios where labeled data is scarce. To address this challenge, researchers have explored Graph Contrastive Learning (GCL), which leverages enhanced graph data and contrastive learning techniques. While promising, existing GCL methods often struggle with effectively capturing both local and global graph structures, and balancing the trade-off between nodelevel and graph-level representations. In this work, we propose Graph Representation Embedding Enhanced via Multidimensional Contrastive Learning (GRE2-MDCL). Our model introduces a novel triple network architecture with a multi-head attention GNN as the core. GRE2-MDCL first globally and locally augments the input graph using SVD and LAGNN techniques. It then constructs a multidimensional contrastive loss, incorporating cross-network, cross-view, and neighbor contrast, to optimize the model. Extensive experiments on benchmark datasets Cora, Citeseer, and PubMed demonstrate that GRE2-MDCL achieves state-of-the-art performance, with average accuracies of 82.5%, 72.5%, and 81.6% respectively. Visualizations further show tighter intra-cluster aggregation and clearer inter-cluster boundaries, highlighting the effectiveness of our framework in improving upon baseline GCL models.

Graph outlier detection is a prominent task of research and application in the realm of graph neural networks. It identifies the outlier nodes that exhibit deviation from the majority in the graph. One of the fundamental challenges confronting supervised graph outlier detection algorithms is the prevalent issue of class imbalance, where the scarcity of outlier instances compared to normal instances often results in suboptimal performance. Conventional methods mitigate the imbalance by reweighting instances in the estimation of the loss function, assigning higher weights to outliers and lower weights to inliers. Nonetheless, these strategies are prone to overfitting and underfitting, respectively. Recently, generative models, especially diffusion models, have demonstrated their efficacy in synthesizing high-fidelity images. Despite their extraordinary generation quality, their potential in data augmentation for supervised graph outlier detection remains largely underexplored. To bridge this gap, we introduce GODM, a novel data augmentation for mitigating class imbalance in supervised Graph Outlier detection with latent Diffusion Models. Specifically, our proposed method consists of three key components: (1) Variantioanl Encoder maps the heterogeneous information inherent within the graph data into a unified latent space. (2) Graph Generator synthesizes graph data that are statistically similar to real outliers from latent space, and (3) Latent Diffusion Model learns the latent space distribution of real organic data by iterative denoising. Extensive experiments conducted on multiple datasets substantiate the effectiveness and efficiency of GODM. The case study further demonstrated the generation quality of our synthetic data. To foster accessibility and reproducibility, we encapsulate GODM into a plug-and-play package and release it at the Python Package Index (PyPI).

Preprint. Under review. Package available at https://pypi.org/project/godm/

E-commerce platforms have a vast catalog of items to cater to their customers' shopping interests. Most of these platforms assist their customers in the shopping process by offering optimized recommendation carousels, designed to help customers quickly locate their desired items. Many models have been proposed in academic literature to generate and enhance the ranking and recall set of items in these carousels. Conventionally, the accompanying carousel title text (header) of these carousels remains static. In most instances, a generic text such as "Items similar to your current viewing" is utilized. Fixed variations such as the inclusion of specific attributes "Other items from a similar seller" or "Items from a similar brand" in addition to "frequently bought together" or "considered together" are observed as well. This work proposes a novel approach to customize the header generation process of these carousels. Our work leverages user-generated reviews that lay focus on specific attributes (aspects) of an item that were favorably perceived by users during their interaction with the given item. We extract these aspects from reviews and train a graph neural network-based model under the framework of a conditional ranking task. We refer to our innovative methodology as Dynamic Text Snippets (DTS) which generates multiple header texts for an anchor item and its recall set. Our approach demonstrates the potential of utilizing user-generated reviews and presents a unique paradigm for exploring increasingly context-aware recommendation systems.

7 pages, 3 figures, PAIS 2024 (ECAI)

Graph Neural Networks (GNNs) have established themselves as a key component in addressing diverse graph-based tasks. Despite their notable successes, GNNs remain susceptible to input perturbations in the form of adversarial attacks. This paper introduces an innovative approach to fortify GNNs against adversarial perturbations through the lens of coupled dynamical systems. Our method introduces graph neural layers based on differential equations with contractive properties, which, as we show, improve the robustness of GNNs. A distinctive feature of the proposed approach is the simultaneous learned evolution of both the node features and the adjacency matrix, yielding an intrinsic enhancement of model robustness to perturbations in the input features and the connectivity of the graph. We mathematically derive the underpinnings of our novel architecture and provide theoretical insights to reason about its expected behavior. We demonstrate the efficacy of our method through numerous real-world benchmarks, reading on par or improved performance compared to existing methods.

Islamophobic language on online platforms fosters intolerance, making detection and elimination crucial for promoting harmony. Traditional hate speech detection models rely on NLP techniques like tokenization, part-of-speech tagging, and encoder-decoder models. However, Graph Neural Networks (GNNs), with their ability to utilize relationships between data points, offer more effective detection and greater explainability. In this work, we represent speeches as nodes and connect them with edges based on their context and similarity to develop the graph. This study introduces a novel paradigm using GNNs to identify and explain hate speech towards Islam. Our model leverages GNNs to understand the context and patterns of hate speech by connecting texts via pretrained NLP-generated word embeddings, achieving state-of-the-art performance and enhancing detection accuracy while providing valuable explanations. This highlights the potential of GNNs in combating online hate speech and fostering a safer, more inclusive online environment.

Accepted in: (i) NeurIPS 2023 : Muslims in ML Workshop (non-archival) (https://www.musiml.org/schedule/#:~:text=Azmine%20Toushik%20Wasi) (ii) EMNLP 2024 : NLP for Positive Impact Workshop (archival)

Neural algorithmic reasoning (NAR) is an emerging field that seeks to design neural networks that mimic classical algorithmic computations. Today, graph neural networks (GNNs) are widely used in neural algorithmic reasoners due to their message passing framework and permutation equivariance. In this extended abstract, we challenge this design choice, and replace the equivariant aggregation function with a recurrent neural network. While seemingly counter-intuitive, this approach has appropriate grounding when nodes have a natural ordering -- and this is the case frequently in established reasoning benchmarks like CLRS-30. Indeed, our recurrent NAR (RNAR) model performs very strongly on such tasks, while handling many others gracefully. A notable achievement of RNAR is its decisive state-of-the-art result on the Heapsort and Quickselect tasks, both deemed as a significant challenge for contemporary neural algorithmic reasoners -- especially the latter, where RNAR achieves a mean micro-F1 score of 87%.

Graphs are an essential data structure utilized to represent relationships in real-world scenarios. Prior research has established that Graph Neural Networks (GNNs) deliver impressive outcomes in graph-centric tasks, such as link prediction and node classification. Despite these advancements, challenges like data sparsity and limited generalization capabilities continue to persist. Recently, Large Language Models (LLMs) have gained attention in natural language processing. They excel in language comprehension and summarization. Integrating LLMs with graph learning techniques has attracted interest as a way to enhance performance in graph learning tasks. In this survey, we conduct an in-depth review of the latest state-of-the-art LLMs applied in graph learning and introduce a novel taxonomy to categorize existing methods based on their framework design. We detail four unique designs: i) GNNs as Prefix, ii) LLMs as Prefix, iii) LLMs-Graphs Integration, and iv) LLMs-Only, highlighting key methodologies within each category. We explore the strengths and limitations of each framework, and emphasize potential avenues for future research, including overcoming current integration challenges between LLMs and graph learning techniques, and venturing into new application areas. This survey aims to serve as a valuable resource for researchers and practitioners eager to leverage large language models in graph learning, and to inspire continued progress in this dynamic field. We consistently maintain the related open-source materials at \url{https://github.com/HKUDS/Awesome-LLM4Graph-Papers}.

Published as a KDD'24 survey paper

Scaling deep learning models has been at the heart of recent revolutions in language modelling and image generation. Practitioners have observed a strong relationship between model size, dataset size, and performance. However, structure-based architectures such as Graph Neural Networks (GNNs) are yet to show the benefits of scale mainly due to the lower efficiency of sparse operations, large data requirements, and lack of clarity about the effectiveness of various architectures. We address this drawback of GNNs by studying their scaling behavior. Specifically, we analyze message-passing networks, graph Transformers, and hybrid architectures on the largest public collection of 2D molecular graphs. For the first time, we observe that GNNs benefit tremendously from the increasing scale of depth, width, number of molecules, number of labels, and the diversity in the pretraining datasets. We further demonstrate strong finetuning scaling behavior on 38 highly competitive downstream tasks, outclassing previous large models. This gives rise to MolGPS, a new graph foundation model that allows to navigate the chemical space, outperforming the previous state-of-the-arts on 26 out the 38 downstream tasks. We hope that our work paves the way for an era where foundational GNNs drive pharmaceutical drug discovery.

Visual robot self-localization is a fundamental problem in visual robot navigation and has been studied across various problem settings, including monocular and sequential localization. However, many existing studies focus primarily on single-robot scenarios, with limited exploration into general settings involving diverse robots connected through wireless networks with constrained communication capacities, such as open-world distributed robot systems. In particular, issues related to the transfer and sharing of key knowledge, such as visual descriptions and visual vocabulary, between robots have been largely neglected. This work introduces a new self-localization framework designed for open-world distributed robot systems that maintains state-of-the-art performance while offering two key advantages: (1) it employs an unsupervised visual vocabulary model that maps to multimodal, lightweight, and transferable visual features, and (2) the visual vocabulary itself is a lightweight and communication-friendly model. Although the primary focus is on encoding monocular view images, the framework can be easily extended to sequential localization applications. By utilizing complementary similarity-preserving features -- both absolute and relative -- the framework meets the requirements for being unsupervised, multimodal, lightweight, and transferable. All features are learned and recognized using a lightweight graph neural network and scene graph. The effectiveness of the proposed method is validated in both passive and active self-localization scenarios.

8 pages, 5 figures, Supplementary Material for International Conference Papers

Heterophilous graphs, where dissimilar nodes tend to connect, pose a challenge for graph neural networks (GNNs) as their superior performance typically comes from aggregating homophilous information. Increasing the GNN depth can expand the scope (i.e., receptive field), potentially finding homophily from the higher-order neighborhoods. However, uniformly expanding the scope results in subpar performance since real-world graphs often exhibit homophily disparity between nodes. An ideal way is personalized scopes, allowing nodes to have varying scope sizes. Existing methods typically add node-adaptive weights for each hop. Although expressive, they inevitably suffer from severe overfitting. To address this issue, we formalize personalized scoping as a separate scope classification problem that overcomes GNN overfitting in node classification. Specifically, we predict the optimal GNN depth for each node. Our theoretical and empirical analysis suggests that accurately predicting the depth can significantly enhance generalization. We further propose Adaptive Scope (AS), a lightweight MLP-based approach that only participates in GNN inference. AS encodes structural patterns and predicts the depth to select the best model for each node's prediction. Experimental results show that AS is highly flexible with various GNN architectures across a wide range of datasets while significantly improving accuracy.

The medical community believes binary medical event outcomes in EHR data contain sufficient information for making a sensible recommendation. However, there are two challenges to effectively utilizing such data: (1) modeling the relationship between massive 0,1 event outcomes is difficult, even with expert knowledge; (2) in practice, learning can be stalled by the binary values since the equally important 0 entries propagate no learning signals. Currently, there is a large gap between the assumed sufficient information and the reality that no promising results have been shown by utilizing solely the binary data: visiting or secondary information is often necessary to reach acceptable performance. In this paper, we attempt to build the first successful binary EHR data-oriented drug recommendation system by tackling the two difficulties, making sensible drug recommendations solely using the binary EHR medical records. To this end, we take a statistical perspective to view the EHR data as a sample from its cohorts and transform them into continuous Bernoulli probabilities. The transformed entries not only model a deterministic binary event with a distribution but also allow reflecting \emph{event-event} relationship by conditional probability. A graph neural network is learned on top of the transformation. It captures event-event correlations while emphasizing \emph{event-to-patient} features. Extensive results demonstrate that the proposed method achieves state-of-the-art performance on large-scale databases, outperforming baseline methods that use secondary information by a large margin. The source code is available at \url{https://github.com/chenzRG/BEHRMecom}

Generative models trained at scale can now produce text, video, and more recently, scientific data such as crystal structures. In applications of generative approaches to materials science, and in particular to crystal structures, the guidance from the domain expert in the form of high-level instructions can be essential for an automated system to output candidate crystals that are viable for downstream research. In this work, we formulate end-to-end language-to-structure generation as a multi-objective optimization problem, and propose Generative Hierarchical Materials Search (GenMS) for controllable generation of crystal structures. GenMS consists of (1) a language model that takes high-level natural language as input and generates intermediate textual information about a crystal (e.g., chemical formulae), and (2) a diffusion model that takes intermediate information as input and generates low-level continuous value crystal structures. GenMS additionally uses a graph neural network to predict properties (e.g., formation energy) from the generated crystal structures. During inference, GenMS leverages all three components to conduct a forward tree search over the space of possible structures. Experiments show that GenMS outperforms other alternatives of directly using language models to generate structures both in satisfying user request and in generating low-energy structures. We confirm that GenMS is able to generate common crystal structures such as double perovskites, or spinels, solely from natural language input, and hence can form the foundation for more complex structure generation in near future.

https://generative-materials.github.io/

Graph neural networks (GNNs) have demonstrated their effectiveness in various tasks supported by their generalization capabilities. However, the current analysis of GNN generalization relies on the assumption that training and testing data are independent and identically distributed (i.i.d). This imposes limitations on the cases where a model mismatch exists when generating testing data. In this paper, we examine GNNs that operate on geometric graphs generated from manifold models, explicitly focusing on scenarios where there is a mismatch between manifold models generating training and testing data. Our analysis reveals the robustness of the GNN generalization in the presence of such model mismatch. This indicates that GNNs trained on graphs generated from a manifold can still generalize well to unseen nodes and graphs generated from a mismatched manifold. We attribute this mismatch to both node feature perturbations and edge perturbations within the generated graph. Our findings indicate that the generalization gap decreases as the number of nodes grows in the training graph while increasing with larger manifold dimension as well as larger mismatch. Importantly, we observe a trade-off between the generalization of GNNs and the capability to discriminate high-frequency components when facing a model mismatch. The most important practical consequence of this analysis is to shed light on the filter design of generalizable GNNs robust to model mismatch. We verify our theoretical findings with experiments on multiple real-world datasets.

20 pages, 6 figures. arXiv admin note: substantial text overlap with arXiv:2406.05225

Convolutional neural networks have been successfully extended to operate on graphs, giving rise to Graph Neural Networks (GNNs). GNNs combine information from adjacent nodes by successive applications of graph convolutions. GNNs have been implemented successfully in various learning tasks while the theoretical understanding of their generalization capability is still in progress. In this paper, we leverage manifold theory to analyze the statistical generalization gap of GNNs operating on graphs constructed on sampled points from manifolds. We study the generalization gaps of GNNs on both node-level and graph-level tasks. We show that the generalization gaps decrease with the number of nodes in the training graphs, which guarantees the generalization of GNNs to unseen points over manifolds. We validate our theoretical results in multiple real-world datasets.

Leveraging network information for predictive modeling has become widespread in many domains. Within the realm of referral and targeted marketing, influencer detection stands out as an area that could greatly benefit from the incorporation of dynamic network representation due to the continuous evolution of customer-brand relationships. In this paper, we present INFLECT-DGNN, a new method for profit-driven INFLuencer prEdiCTion with Dynamic Graph Neural Networks that innovatively combines Graph Neural Networks (GNNs) and Recurrent Neural Networks (RNNs) with weighted loss functions, synthetic minority oversampling adapted to graph data, and a carefully crafted rolling-window strategy. We introduce a novel profit-driven framework that supports decision-making based on model predictions. To test the framework, we use a unique corporate dataset with diverse networks, capturing the customer interactions across three cities with different socioeconomic and demographic characteristics. Our results show how using RNNs to encode temporal attributes alongside GNNs significantly improves predictive performance, while the profit-driven framework determines the optimal classification threshold for profit maximization. We compare the results of different models to demonstrate the importance of capturing network representation, temporal dependencies, and using a profit-driven evaluation. Our research has significant implications for the fields of referral and targeted marketing, expanding the technical use of deep graph learning within corporate environments.

Image analysis in the euclidean space through linear hyperspaces is well studied. However, in the quest for more effective image representations, we turn to hyperbolic manifolds. They provide a compelling alternative to capture complex hierarchical relationships in images with remarkably small dimensionality. To demonstrate hyperbolic embeddings' competence, we introduce a light-weight hyperbolic graph neural network for image segmentation, encompassing patch-level features in a very small embedding size. Our solution, Seg-HGNN, surpasses the current best unsupervised method by 2.5\%, 4\% on VOC-07, VOC-12 for localization, and by 0.8\%, 1.3\% on CUB-200, ECSSD for segmentation, respectively. With less than 7.5k trainable parameters, Seg-HGNN delivers effective and fast ($\approx 2$ images/second) results on very standard GPUs like the GTX1650. This empirical evaluation presents compelling evidence of the efficacy and potential of hyperbolic representations for vision tasks.

We present $\textit{Learn2Aggregate}$, a machine learning (ML) framework for optimizing the generation of Chv\'atal-Gomory (CG) cuts in mixed integer linear programming (MILP). The framework trains a graph neural network to classify useful constraints for aggregation in CG cut generation. The ML-driven CG separator selectively focuses on a small set of impactful constraints, improving runtimes without compromising the strength of the generated cuts. Key to our approach is the formulation of a constraint classification task which favours sparse aggregation of constraints, consistent with empirical findings. This, in conjunction with a careful constraint labeling scheme and a hybrid of deep learning and feature engineering, results in enhanced CG cut generation across five diverse MILP benchmarks. On the largest test sets, our method closes roughly $\textit{twice}$ as much of the integrality gap as the standard CG method while running 40$% faster. This performance improvement is due to our method eliminating 75% of the constraints prior to aggregation.

The discrete Laplacian operator holds a crucial role in 3D geometry processing, yet it is still challenging to define it on point clouds. Previous works mainly focused on constructing a local triangulation around each point to approximate the underlying manifold for defining the Laplacian operator, which may not be robust or accurate. In contrast, we simply use the K-nearest neighbors (KNN) graph constructed from the input point cloud and learn the Laplacian operator on the KNN graph with graph neural networks (GNNs). However, the ground-truth Laplacian operator is defined on a manifold mesh with a different connectivity from the KNN graph and thus cannot be directly used for training. To train the GNN, we propose a novel training scheme by imitating the behavior of the ground-truth Laplacian operator on a set of probe functions so that the learned Laplacian operator behaves similarly to the ground-truth Laplacian operator. We train our network on a subset of ShapeNet and evaluate it across a variety of point clouds. Compared with previous methods, our method reduces the error by an order of magnitude and excels in handling sparse point clouds with thin structures or sharp features. Our method also demonstrates a strong generalization ability to unseen shapes. With our learned Laplacian operator, we further apply a series of Laplacian-based geometry processing algorithms directly to point clouds and achieve accurate results, enabling many exciting possibilities for geometry processing on point clouds. The code and trained models are available at https://github.com/IntelligentGeometry/NeLo.

SIGGRAPH Asia 2024 (Journal Track)

Spatio-temporal prediction is a crucial research area in data-driven urban computing, with implications for transportation, public safety, and environmental monitoring. However, scalability and generalization challenges remain significant obstacles. Advanced models often rely on Graph Neural Networks to encode spatial and temporal correlations, but struggle with the increased complexity of large-scale datasets. The recursive GNN-based message passing schemes used in these models hinder their training and deployment in real-life urban sensing scenarios. Moreover, long-spanning large-scale spatio-temporal data introduce distribution shifts, necessitating improved generalization performance. To address these challenges, we propose a simple framework for spatio-temporal prediction - EasyST paradigm. It learns lightweight and robust Multi-Layer Perceptrons (MLPs) by effectively distilling knowledge from complex spatio-temporal GNNs. We ensure robust knowledge distillation by integrating the spatio-temporal information bottleneck with teacher-bounded regression loss, filtering out task-irrelevant noise and avoiding erroneous guidance. We further enhance the generalization ability of the student model by incorporating spatial and temporal prompts to provide downstream task contexts. Evaluation on three spatio-temporal datasets for urban computing tasks demonstrates that EasyST surpasses state-of-the-art approaches in terms of efficiency and accuracy. The implementation code is available at: https://github.com/HKUDS/EasyST.

Accepted by CIKM'2024, full paper

Graph Neural Network (GNN) models on streaming graphs entail algorithmic challenges to continuously capture its dynamic state, as well as systems challenges to optimize latency, memory, and throughput during both inference and training. We present D3-GNN, the first distributed, hybrid-parallel, streaming GNN system designed to handle real-time graph updates under online query setting. Our system addresses data management, algorithmic, and systems challenges, enabling continuous capturing of the dynamic state of the graph and updating node representations with fault-tolerance and optimal latency, load-balance, and throughput. D3-GNN utilizes streaming GNN aggregators and an unrolled, distributed computation graph architecture to handle cascading graph updates. To counteract data skew and neighborhood explosion issues, we introduce inter-layer and intra-layer windowed forward pass solutions. Experiments on large-scale graph streams demonstrate that D3-GNN achieves high efficiency and scalability. Compared to DGL, D3-GNN achieves a significant throughput improvement of about 76x for streaming workloads. The windowed enhancement further reduces running times by around 10x and message volumes by up to 15x at higher parallelism.

14 pages, 7 figures, published at VLDB'24

Social Event Detection (SED) aims to identify significant events from social streams, and has a wide application ranging from public opinion analysis to risk management. In recent years, Graph Neural Network (GNN) based solutions have achieved state-of-the-art performance. However, GNN-based methods often struggle with missing and noisy edges between messages, affecting the quality of learned message embedding. Moreover, these methods statically initialize node embedding before training, which, in turn, limits the ability to learn from message texts and relations simultaneously. In this paper, we approach social event detection from a new perspective based on Pre-trained Language Models (PLMs), and present RPLM_SED (Relational prompt-based Pre-trained Language Models for Social Event Detection). We first propose a new pairwise message modeling strategy to construct social messages into message pairs with multi-relational sequences. Secondly, a new multi-relational prompt-based pairwise message learning mechanism is proposed to learn more comprehensive message representation from message pairs with multi-relational prompts using PLMs. Thirdly, we design a new clustering constraint to optimize the encoding process by enhancing intra-cluster compactness and inter-cluster dispersion, making the message representation more distinguishable. We evaluate the RPLM_SED on three real-world datasets, demonstrating that the RPLM_SED model achieves state-of-the-art performance in offline, online, low-resource, and long-tail distribution scenarios for social event detection tasks.

Heterogeneous graph neural networks (HGNNs) have significantly propelled the information retrieval (IR) field. Still, the effectiveness of HGNNs heavily relies on high-quality labels, which are often expensive to acquire. This challenge has shifted attention towards Heterogeneous Graph Contrastive Learning (HGCL), which usually requires pre-defined meta-paths. However, our findings reveal that meta-path combinations significantly affect performance in unsupervised settings, an aspect often overlooked in current literature. Existing HGCL methods have considerable variability in outcomes across different meta-path combinations, thereby challenging the optimization process to achieve consistent and high performance. In response, we introduce \textsf{LAMP} (\underline{\textbf{L}}earn\underline{\textbf{A}}ble \underline{\textbf{M}}eta-\underline{\textbf{P}}ath), a novel adversarial contrastive learning approach that integrates various meta-path sub-graphs into a unified and stable structure, leveraging the overlap among these sub-graphs. To address the denseness of this integrated sub-graph, we propose an adversarial training strategy for edge pruning, maintaining sparsity to enhance model performance and robustness. \textsf{LAMP} aims to maximize the difference between meta-path and network schema views for guiding contrastive learning to capture the most meaningful information. Our extensive experimental study conducted on four diverse datasets from the Heterogeneous Graph Benchmark (HGB) demonstrates that \textsf{LAMP} significantly outperforms existing state-of-the-art unsupervised models in terms of accuracy and robustness.

We propose a novel approach to compute the MAXCUT in attributed graphs, i.e., graphs with features associated with nodes and edges. Our approach is robust to the underlying graph topology and is fully differentiable, making it possible to find solutions that jointly optimize the MAXCUT along with other objectives. Based on the obtained MAXCUT partition, we implement a hierarchical graph pooling layer for Graph Neural Networks, which is sparse, differentiable, and particularly suitable for downstream tasks on heterophilic graphs.

Multitask learning is a widely used paradigm for training models on diverse tasks, with applications ranging from graph neural networks to language model fine-tuning. Since tasks may interfere with each other, a key notion for modeling their relationships is task affinity. This includes pairwise task affinity, computed among pairs of tasks, and higher-order affinity, computed among subsets of tasks. Naively computing either of them requires repeatedly training on data from various task combinations, which is computationally intensive. We present a new algorithm Grad-TAG that can estimate task affinities without this repeated training. The key idea of Grad-TAG is to train a "base" model for all tasks and then use a linearization technique to estimate the loss of the model for a specific task combination. The linearization works by computing a gradient-based approximation of the loss, using low-dimensional projections of gradients as features in a logistic regression to predict labels for the task combination. We show that the linearized model can provably approximate the loss when the gradient-based approximation is accurate, and also empirically verify that on several large models. Then, given the estimated task affinity, we design a semi-definite program for clustering similar tasks by maximizing the average density of clusters. We evaluate Grad-TAG's performance across seven datasets, including multi-label classification on graphs, and instruction fine-tuning of language models. Our task affinity estimates are within 2.7% distance to the true affinities while needing only 3% of FLOPs in full training. On our largest graph with 21M edges and 500 labeling tasks, our algorithm delivers estimates within 5% distance to the true affinities, using only 112 GPU hours. Our results show that Grad-TAG achieves excellent performance and runtime tradeoffs compared to existing approaches.

Precise and timely fault diagnosis is a prerequisite for a distribution system to ensure minimum downtime and maintain reliable operation. This necessitates access to a comprehensive procedure that can provide the grid operators with insightful information in the case of a fault event. In this paper, we propose a heterogeneous multi-task learning graph neural network (MTL-GNN) capable of detecting, locating and classifying faults in addition to providing an estimate of the fault resistance and current. Using a graph neural network (GNN) allows for learning the topological representation of the distribution system as well as feature learning through a message-passing scheme. We investigate the robustness of our proposed model using the IEEE-123 test feeder system. This work also proposes a novel GNN-based explainability method to identify key nodes in the distribution system which then facilitates informed sparse measurements. Numerical tests validate the performance of the model across all tasks.

Published in IEEE Transactions on Power Systems (2024)

Graph Neural Networks (GNNs) have emerged as powerful tools for learning representations of graph-structured data, demonstrating remarkable performance across various tasks. Recognising their importance, there has been extensive research focused on explaining GNN predictions, aiming to enhance their interpretability and trustworthiness. However, GNNs and their explainers face a notable challenge: graphs are primarily designed to model pair-wise relationships between nodes, which can make it tough to capture higher-order, multi-node interactions. This characteristic can pose difficulties for existing explainers in fully representing multi-node relationships. To address this gap, we present Framework For Higher-Order Representations In Graph Explanations (FORGE), a framework that enables graph explainers to capture such interactions by incorporating higher-order structures, resulting in more accurate and faithful explanations. Extensive evaluation shows that on average real-world datasets from the GraphXAI benchmark and synthetic datasets across various graph explainers, FORGE improves average explanation accuracy by 1.9x and 2.25x, respectively. We perform ablation studies to confirm the importance of higher-order relations in improving explanations, while our scalability analysis demonstrates FORGE's efficacy on large graphs.

Celcomen leverages a mathematical causality framework to disentangle intra- and inter- cellular gene regulation programs in spatial transcriptomics and single-cell data through a generative graph neural network. It can learn gene-gene interactions, as well as generate post-perturbation counterfactual spatial transcriptomics, thereby offering access to experimentally inaccessible samples. We validated its disentanglement, identifiability, and counterfactual prediction capabilities through simulations and in clinically relevant human glioblastoma, human fetal spleen, and mouse lung cancer samples. Celcomen provides the means to model disease and therapy induced changes allowing for new insights into single-cell spatially resolved tissue responses relevant to human health.

Over the past decade, Graph Neural Networks (GNNs) have achieved great success on machine learning tasks with relational data. However, recent studies have found that heterophily can cause significant performance degradation of GNNs, especially on node-level tasks. Numerous heterophilic benchmark datasets have been put forward to validate the efficacy of heterophily-specific GNNs and various homophily metrics have been designed to help people recognize these malignant datasets. Nevertheless, there still exist multiple pitfalls that severely hinder the proper evaluation of new models and metrics. In this paper, we point out three most serious pitfalls: 1) a lack of hyperparameter tuning; 2) insufficient model evaluation on the real challenging heterophilic datasets; 3) missing quantitative evaluation benchmark for homophily metrics on synthetic graphs. To overcome these challenges, we first train and fine-tune baseline models on $27$ most widely used benchmark datasets, categorize them into three distinct groups: malignant, benign and ambiguous heterophilic datasets, and identify the real challenging subsets of tasks. To our best knowledge, we are the first to propose such taxonomy. Then, we re-evaluate $10$ heterophily-specific state-of-the-arts (SOTA) GNNs with fine-tuned hyperparameters on different groups of heterophilic datasets. Based on the model performance, we reassess their effectiveness on addressing heterophily challenge. At last, we evaluate $11$ popular homophily metrics on synthetic graphs with three different generation approaches. To compare the metrics strictly, we propose the first quantitative evaluation method based on Fr\'echet distance.

arXiv admin note: substantial text overlap with arXiv:2407.09618

Graph neural network (GNN) has been a powerful approach in collaborative filtering (CF) due to its ability to model high-order user-item relationships. Recently, to alleviate the data sparsity and enhance representation learning, many efforts have been conducted to integrate contrastive learning (CL) with GNNs. Despite the promising improvements, the contrastive view generation based on structure and representation perturbations in existing methods potentially disrupts the collaborative information in contrastive views, resulting in limited effectiveness of positive alignment. To overcome this issue, we propose CoGCL, a novel framework that aims to enhance graph contrastive learning by constructing contrastive views with stronger collaborative information via discrete codes. The core idea is to map users and items into discrete codes rich in collaborative information for reliable and informative contrastive view generation. To this end, we initially introduce a multi-level vector quantizer in an end-to-end manner to quantize user and item representations into discrete codes. Based on these discrete codes, we enhance the collaborative information of contrastive views by considering neighborhood structure and semantic relevance respectively. For neighborhood structure, we propose virtual neighbor augmentation by treating discrete codes as virtual neighbors, which expands an observed user-item interaction into multiple edges involving discrete codes. Regarding semantic relevance, we identify similar users/items based on shared discrete codes and interaction targets to generate the semantically relevant view. Through these strategies, we construct contrastive views with stronger collaborative information and develop a triple-view graph contrastive learning approach. Extensive experiments on four public datasets demonstrate the effectiveness of our proposed approach.

Recent years have witnessed great success in handling graph-related tasks with Graph Neural Networks (GNNs). However, most existing GNNs are based on message passing to perform feature aggregation and transformation, where the structural information is explicitly involved in the forward propagation by coupling with node features through graph convolution at each layer. As a result, subtle feature noise or structure perturbation may cause severe error propagation, resulting in extremely poor robustness. In this paper, we rethink the roles played by graph structural information in graph data training and identify that message passing is not the only path to modeling structural information. Inspired by this, we propose a simple but effective Graph Structure Self-Contrasting (GSSC) framework that learns graph structural information without message passing. The proposed framework is based purely on Multi-Layer Perceptrons (MLPs), where the structural information is only implicitly incorporated as prior knowledge to guide the computation of supervision signals, substituting the explicit message propagation as in GNNs. Specifically, it first applies structural sparsification to remove potentially uninformative or noisy edges in the neighborhood, and then performs structural self-contrasting in the sparsified neighborhood to learn robust node representations. Finally, structural sparsification and self-contrasting are formulated as a bi-level optimization problem and solved in a unified framework. Extensive experiments have qualitatively and quantitatively demonstrated that the GSSC framework can produce truly encouraging performance with better generalization and robustness than other leading competitors.

Since the onset of the COVID-19 pandemic, there has been a growing interest in studying epidemiological models. Traditional mechanistic models mathematically describe the transmission mechanisms of infectious diseases. However, they often suffer from limitations of oversimplified or fixed assumptions, which could cause sub-optimal predictive power and inefficiency in capturing complex relation information. Consequently, Graph Neural Networks(GNNs) have emerged as a progressively popular tool in epidemic research. In this paper, we endeavor to furnish a comprehensive review of GNNs in epidemic tasks and highlight potential future directions. To accomplish this objective, we introduce hierarchical taxonomies for both epidemic tasks and methodologies, offering a trajectory of development within this domain. For epidemic tasks, we establish a taxonomy akin to those typically employed within the epidemic domain. For methodology, we categorize existing work into Neural Models and Hybrid Models. Following this, we perform an exhaustive and systematic examination of the methodologies, encompassing both the tasks and their technical details. Furthermore, we discuss the limitations of existing methods from diverse perspectives and systematically propose future research directions. This survey aims to bridge literature gaps and promote the progression of this promising field, with a list of relevant papers at https://github.com/Emory-Melody/awesome-epidemic-modeling-papers. We hope that it will facilitate synergies between the communities of GNNs and epidemiology, and contribute to their collective progress.

The great success of Physics-Informed Neural Networks (PINN) in solving partial differential equations (PDEs) has significantly advanced our simulation and understanding of complex physical systems in science and engineering. However, many PINN-like methods are poorly scalable and are limited to in-sample scenarios. To address these challenges, this work proposes a novel discrete approach termed Physics-Informed Graph Neural Network (PIGNN) to solve forward and inverse nonlinear PDEs. In particular, our approach seamlessly integrates the strength of graph neural networks (GNN), physical equations and finite difference to approximate solutions of physical systems. Our approach is compared with the PINN baseline on three well-known nonlinear PDEs (heat, Burgers and FitzHugh-Nagumo). We demonstrate the excellent performance of the proposed method to work with irregular meshes, longer time steps, arbitrary spatial resolutions, varying initial conditions (ICs) and boundary conditions (BCs) by conducting extensive numerical experiments. Numerical results also illustrate the superiority of our approach in terms of accuracy, time extrapolability, generalizability and scalability. The main advantage of our approach is that models trained in small domains with simple settings have excellent fitting capabilities and can be directly applied to more complex situations in large domains.

Ethereum faces growing fraud threats. Current fraud detection methods, whether employing graph neural networks or sequence models, fail to consider the semantic information and similarity patterns within transactions. Moreover, these approaches do not leverage the potential synergistic benefits of combining both types of models. To address these challenges, we propose TLMG4Eth that combines a transaction language model with graph-based methods to capture semantic, similarity, and structural features of transaction data in Ethereum. We first propose a transaction language model that converts numerical transaction data into meaningful transaction sentences, enabling the model to learn explicit transaction semantics. Then, we propose a transaction attribute similarity graph to learn transaction similarity information, enabling us to capture intuitive insights into transaction anomalies. Additionally, we construct an account interaction graph to capture the structural information of the account transaction network. We employ a deep multi-head attention network to fuse transaction semantic and similarity embeddings, and ultimately propose a joint training approach for the multi-head attention network and the account interaction graph to obtain the synergistic benefits of both.

Graph representation learning (GRL) models have succeeded in many scenarios. Real-world graphs have imbalanced distribution, such as node labels and degrees, which leaves a critical challenge to GRL. Imbalanced inputs can lead to imbalanced outputs. However, most existing works ignore it and assume that the distribution of input graphs is balanced, which cannot align with real situations, resulting in worse model performance on tail data. The domination of head data makes tail data underrepresented when training graph neural networks (GNNs). Thus, we propose Graffin, a pluggable tail data augmentation module, to address the above issues. Inspired by recurrent neural networks (RNNs), Graffin flows head features into tail data through graph serialization techniques to alleviate the imbalance of tail representation. The local and global structures are fused to form the node representation under the combined effect of neighborhood and sequence information, which enriches the semantics of tail data. We validate the performance of Graffin on four real-world datasets in node classification tasks. Results show that Graffin can improve the adaptation to tail data without significantly degrading the overall model performance.

Online conversations are particularly susceptible to derailment, which can manifest itself in the form of toxic communication patterns including disrespectful comments and abuse. Forecasting conversation derailment predicts signs of derailment in advance enabling proactive moderation of conversations. State-of-the-art approaches to conversation derailment forecasting sequentially encode conversations and use graph neural networks to model dialogue user dynamics. However, existing graph models are not able to capture complex conversational characteristics such as context propagation and emotional shifts. The use of common sense knowledge enables a model to capture such characteristics, thus improving performance. Following this approach, here we derive commonsense statements from a knowledge base of dialogue contextual information to enrich a graph neural network classification architecture. We fuse the multi-source information on utterance into capsules, which are used by a transformer-based forecaster to predict conversation derailment. Our model captures conversation dynamics and context propagation, outperforming the state-of-the-art models on the CGA and CMV benchmark datasets

arXiv admin note: substantial text overlap with arXiv:2306.12982; text overlap with arXiv:2106.01071 by other authors

In this paper, we study the generalization capabilities of geometric graph neural networks (GNNs). We consider GNNs over a geometric graph constructed from a finite set of randomly sampled points over an embedded manifold with topological information captured. We prove a generalization gap between the optimal empirical risk and the optimal statistical risk of this GNN, which decreases with the number of sampled points from the manifold and increases with the dimension of the underlying manifold. This generalization gap ensures that the GNN trained on a graph on a set of sampled points can be utilized to process other unseen graphs constructed from the same underlying manifold. The most important observation is that the generalization capability can be realized with one large graph instead of being limited to the size of the graph as in previous results. The generalization gap is derived based on the non-asymptotic convergence result of a GNN on the sampled graph to the underlying manifold neural networks (MNNs). We verify this theoretical result with experiments on both Arxiv dataset and Cora dataset.

12 pages, 4 figures. arXiv admin note: text overlap with arXiv:2406.05225

Depth estimation plays a pivotal role in autonomous driving, facilitating a comprehensive understanding of the vehicle's 3D surroundings. Radar, with its robustness to adverse weather conditions and capability to measure distances, has drawn significant interest for radar-camera depth estimation. However, existing algorithms process the inherently noisy and sparse radar data by projecting 3D points onto the image plane for pixel-level feature extraction, overlooking the valuable geometric information contained within the radar point cloud. To address this gap, we propose GET-UP, leveraging attention-enhanced Graph Neural Networks (GNN) to exchange and aggregate both 2D and 3D information from radar data. This approach effectively enriches the feature representation by incorporating spatial relationships compared to traditional methods that rely only on 2D feature extraction. Furthermore, we incorporate a point cloud upsampling task to densify the radar point cloud, rectify point positions, and derive additional 3D features under the guidance of lidar data. Finally, we fuse radar and camera features during the decoding phase for depth estimation. We benchmark our proposed GET-UP on the nuScenes dataset, achieving state-of-the-art performance with a 15.3% and 14.7% improvement in MAE and RMSE over the previously best-performing model. Code: https://github.com/harborsarah/GET-UP

Accepted by WACV 2025

Polycrystal plasticity in metals is characterized by nonlinear behavior and strain hardening, making numerical models computationally intensive. We employ Graph Neural Network (GNN) to surrogate polycrystal plasticity from finite element method (FEM) simulations. We present a novel message-passing GNN that encodes nodal strain and edge distances between FEM mesh cells, aggregates them to obtain embeddings, and combines the decoded embeddings with the nodal strains to predict stress tensors on graph nodes. We demonstrate training GNN based on subgraphs generated from FEM mesh-graphs, in which the mesh cells are converted to nodes and edges are created between adjacent cells. The GNN is trained on 72 graphs and tested on 18 graphs. We apply the trained GNN to periodic polycrystals and learn the stress-strain maps based on strain-gradient plasticity theory. The GNN is accurately trained based on FEM graphs, in which the $R^2$ for both training and testing sets are 0.993. The proposed GNN plasticity constitutive model speeds up more than 150 times compared with the benchmark FEM method on randomly selected test polycrystals. We also apply the trained GNN to 30 unseen FEM simulations and the GNN generalizes well with an overall $R^2$ of 0.992. Analysis of the von Mises stress distributions in polycrystals shows that the GNN model accurately learns the stress distribution with low error. By comparing the error distribution across training, testing, and unseen datasets, we can deduce that the proposed model does not overfit and generalizes well beyond the training data. This work is expected to pave the way for using graphs as surrogates in polycrystal plasticity modeling.

Document-level relation extraction (DocRE) aims to extract relations between entities from unstructured document text. Compared to sentence-level relation extraction, it requires more complex semantic understanding from a broader text context. Currently, some studies are utilizing logical rules within evidence sentences to enhance the performance of DocRE. However, in the data without provided evidence sentences, researchers often obtain a list of evidence sentences for the entire document through evidence retrieval (ER). Therefore, DocRE suffers from two challenges: firstly, the relevance between evidence and entity pairs is weak; secondly, there is insufficient extraction of complex cross-relations between long-distance multi-entities. To overcome these challenges, we propose GEGA, a novel model for DocRE. The model leverages graph neural networks to construct multiple weight matrices, guiding attention allocation to evidence sentences. It also employs multi-scale representation aggregation to enhance ER. Subsequently, we integrate the most efficient evidence information to implement both fully supervised and weakly supervised training processes for the model. We evaluate the GEGA model on three widely used benchmark datasets: DocRED, Re-DocRED, and Revisit-DocRED. The experimental results indicate that our model has achieved comprehensive improvements compared to the existing SOTA model.

Deep reinforcement learning (DRL) methods that incorporate graph neural networks (GNNs) have been extensively studied for intelligent traffic signal control, which aims to coordinate traffic signals effectively across multiple intersections. Despite this progress, the standard graph learning used in these methods still struggles to capture higher-order correlations in real-world traffic flow. In this paper, we propose a multi-agent proximal policy optimization framework DHG-PPO, which incorporates PPO and directed hypergraph module to extract the spatio-temporal attributes of the road networks. DHG-PPO enables multiple agents to ingeniously interact through the dynamical construction of hypergraph. The effectiveness of DHG-PPO is validated in terms of average travel time and throughput against state-of-the-art baselines through extensive experiments.