Abstract
Denoising Diffusion Probabilistic Model (DDPM) is able to make flexible conditional image generation from prior noise to real data, by introducing an independent noise-aware classifier to provide conditional gradient guidance at each time step of denoising process. However, due to the ability of classifier to easily discriminate an incompletely generated image only with high-level structure, the gradient, which is a kind of class information guidance, tends to vanish early, leading to the collapse from conditional generation process into the unconditional process. To address this problem, we propose two simple but effective approaches from two perspectives. For sampling procedure, we introduce the entropy of predicted distribution as the measure of guidance vanishing level and propose an entropy-aware scaling method to adaptively recover the conditional semantic guidance. % for each generated sample. For training stage, we propose the entropy-aware optimization objectives to alleviate the overconfident prediction for noisy data.On ImageNet1000 256x256, with our proposed sampling scheme and trained classifier, the pretrained conditional and unconditional DDPM model can achieve 10.89% (4.59 to 4.09) and 43.5% (12 to 6.78) FID improvement respectively.
Keyword: overconfidence
There is no result
Keyword: confidence
Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation
Abstract
We study reinforcement learning with linear function approximation where the transition probability and reward functions are linear with respect to a feature mapping $\boldsymbol{\phi}(s,a)$. Specifically, we consider the episodic inhomogeneous linear Markov Decision Process (MDP), and propose a novel computation-efficient algorithm, LSVI-UCB$^+$, which achieves an $\widetilde{O}(Hd\sqrt{T})$ regret bound where $H$ is the episode length, $d$ is the feature dimension, and $T$ is the number of steps. LSVI-UCB$^+$ builds on weighted ridge regression and upper confidence value iteration with a Bernstein-type exploration bonus. Our statistical results are obtained with novel analytical tools, including a new Bernstein self-normalized bound with conservatism on elliptical potentials, and refined analysis of the correction term. To the best of our knowledge, this is the first minimax optimal algorithm for linear MDPs up to logarithmic factors, which closes the $\sqrt{Hd}$ gap between the best known upper bound of $\widetilde{O}(\sqrt{H^3d^3T})$ in \cite{jin2020provably} and lower bound of $\Omega(Hd\sqrt{T})$ for linear MDPs.
A Geometric Method for Improved Uncertainty Estimation in Real-time
Authors: Gabriella Chouraqui, Liron Cohen, Gil Einziger, Liel Leman
Abstract
Machine learning classifiers are probabilistic in nature, and thus inevitably involve uncertainty. Predicting the probability of a specific input to be correct is called uncertainty (or confidence) estimation and is crucial for risk management. Post-hoc model calibrations can improve models' uncertainty estimations without the need for retraining, and without changing the model. Our work puts forward a geometric-based approach for uncertainty estimation. Roughly speaking, we use the geometric distance of the current input from the existing training inputs as a signal for estimating uncertainty and then calibrate that signal (instead of the model's estimation) using standard post-hoc calibration techniques. We show that our method yields better uncertainty estimations than recently proposed approaches by extensively evaluating multiple datasets and models. In addition, we also demonstrate the possibility of performing our approach in near real-time applications. Our code is available at our Github https://github.com/NoSleepDeveloper/Geometric-Calibrator.
Keyword: scaling
Entropy-driven Sampling and Training Scheme for Conditional Diffusion Generation
Authors: Shengming Li, Guangcong Zheng, Hui Wang, Taiping Yao, Yang Chen, Shoudong Ding, Xi Li
Subjects: Computer Vision and Pattern Recognition (cs.CV)
Abstract
Denoising Diffusion Probabilistic Model (DDPM) is able to make flexible conditional image generation from prior noise to real data, by introducing an independent noise-aware classifier to provide conditional gradient guidance at each time step of denoising process. However, due to the ability of classifier to easily discriminate an incompletely generated image only with high-level structure, the gradient, which is a kind of class information guidance, tends to vanish early, leading to the collapse from conditional generation process into the unconditional process. To address this problem, we propose two simple but effective approaches from two perspectives. For sampling procedure, we introduce the entropy of predicted distribution as the measure of guidance vanishing level and propose an entropy-aware scaling method to adaptively recover the conditional semantic guidance. % for each generated sample. For training stage, we propose the entropy-aware optimization objectives to alleviate the overconfident prediction for noisy data.On ImageNet1000 256x256, with our proposed sampling scheme and trained classifier, the pretrained conditional and unconditional DDPM model can achieve 10.89% (4.59 to 4.09) and 43.5% (12 to 6.78) FID improvement respectively.
Abstract
In most of the literature on federated learning (FL), neural networks are initialized with random weights. In this paper, we present an empirical study on the effect of pre-training on FL. Specifically, we aim to investigate if pre-training can alleviate the drastic accuracy drop when clients' decentralized data are non-IID. We focus on FedAvg, the fundamental and most widely used FL algorithm. We found that pre-training does largely close the gap between FedAvg and centralized learning under non-IID data, but this does not come from alleviating the well-known model drifting problem in FedAvg's local training. Instead, how pre-training helps FedAvg is by making FedAvg's global aggregation more stable. When pre-training using real data is not feasible for FL, we propose a novel approach to pre-train with synthetic data. On various image datasets (including one for segmentation), our approach with synthetic pre-training leads to a notable gain, essentially a critical step toward scaling up federated learning for real-world applications.
A Survey on Learnable Evolutionary Algorithms for Scalable Multiobjective Optimization
Authors: Songbai Liu
Subjects: Neural and Evolutionary Computing (cs.NE); Artificial Intelligence (cs.AI)
Abstract
Recent decades have witnessed remarkable advancements in multiobjective evolutionary algorithms (MOEAs) that have been adopted to solve various multiobjective optimization problems (MOPs). However, these progressively improved MOEAs have not necessarily been equipped with sophisticatedly scalable and learnable problem-solving strategies that are able to cope with new and grand challenges brought by the scaling-up MOPs with continuously increasing complexity or scale from diverse aspects, mainly including expensive function evaluations, many objectives, large-scale search space, time-varying environments, and multitask. Under different scenarios, it requires divergent thinking to design new powerful MOEAs for solving them effectively. In this context, research into learnable MOEAs that arm themselves with machine learning techniques for scaling-up MOPs has received extensive attention in the field of evolutionary computation. In this paper, we begin with a taxonomy of scalable MOPs and learnable MOEAs, followed by an analysis of the challenges that scaling up MOPs pose to traditional MOEAs. Then, we synthetically overview recent advances of learnable MOEAs in solving various scaling up MOPs, focusing primarily on three attractive and promising directions (i.e., learnable evolutionary discriminators for environmental selection, learnable evolutionary generators for reproduction, and learnable evolutionary transfer for sharing or reusing optimization experience between different problem domains). The insight into learnable MOEAs held throughout this paper is offered to the readers as a reference to the general track of the efforts in this field.
INTERPLAY: An Intelligent Model for Predicting Performance Degradation due to Multi-cache Way-disabling
Authors: Panagiota Nikolaou, Yiannakis Sazeides, Maria K. Michael
Abstract
Modern and future processors need to remain functionally correct in the presence of permanent faults to sustain scaling benefits and limit field returns. This paper presents a combined analytical and microarchitectural simulation-based framework called INTERPLAY, that can rapidly predict, at design-time, the performance degradation expected from a processor employing way-disabling to handle permanent faults in caches while in-the-field. The proposed model can predict a program's performance with an accuracy of up to 98.40% for a processor with a two-level cache hierarchy, when multiple caches suffer from faults and need to disable one or more of their ways. INTERPLAY is 9.2x faster than an exhaustive simulation approach since it only needs the training simulation runs for the single-cache way-disabling configurations to predict the performance for any multi-cache way-disabling configuration.
Keyword: calibration
Modular Conformal Calibration
Authors: Charles Marx, Shengjia Zhou, Willie Neiswanger, Stefano Ermon
Abstract
Uncertainty estimates must be calibrated (i.e., accurate) and sharp (i.e., informative) in order to be useful. This has motivated a variety of methods for recalibration, which use held-out data to turn an uncalibrated model into a calibrated model. However, the applicability of existing methods is limited due to their assumption that the original model is also a probabilistic model. We introduce a versatile class of algorithms for recalibration in regression that we call Modular Conformal Calibration (MCC). This framework allows one to transform any regression model into a calibrated probabilistic model. The modular design of MCC allows us to make simple adjustments to existing algorithms that enable well-behaved distribution predictions. We also provide finite-sample calibration guarantees for MCC algorithms. Our framework recovers isotonic recalibration, conformal calibration, and conformal interval prediction, implying that our theoretical results apply to those methods as well. Finally, we conduct an empirical study of MCC on 17 regression datasets. Our results show that new algorithms designed in our framework achieve near-perfect calibration and improve sharpness relative to existing methods.
A Geometric Method for Improved Uncertainty Estimation in Real-time
Authors: Gabriella Chouraqui, Liron Cohen, Gil Einziger, Liel Leman
Abstract
Machine learning classifiers are probabilistic in nature, and thus inevitably involve uncertainty. Predicting the probability of a specific input to be correct is called uncertainty (or confidence) estimation and is crucial for risk management. Post-hoc model calibrations can improve models' uncertainty estimations without the need for retraining, and without changing the model. Our work puts forward a geometric-based approach for uncertainty estimation. Roughly speaking, we use the geometric distance of the current input from the existing training inputs as a signal for estimating uncertainty and then calibrate that signal (instead of the model's estimation) using standard post-hoc calibration techniques. We show that our method yields better uncertainty estimations than recently proposed approaches by extensively evaluating multiple datasets and models. In addition, we also demonstrate the possibility of performing our approach in near real-time applications. Our code is available at our Github https://github.com/NoSleepDeveloper/Geometric-Calibrator.
Keyword: out of distribution detection
There is no result
Keyword: out-of-distribution detection
There is no result
Keyword: expected calibration error
There is no result
Keyword: overconfident
Entropy-driven Sampling and Training Scheme for Conditional Diffusion Generation
Keyword: overconfidence
There is no result
Keyword: confidence
Nearly Minimax Optimal Reinforcement Learning with Linear Function Approximation
A Geometric Method for Improved Uncertainty Estimation in Real-time
Keyword: scaling
Entropy-driven Sampling and Training Scheme for Conditional Diffusion Generation
On Pre-Training for Federated Learning
A Survey on Learnable Evolutionary Algorithms for Scalable Multiobjective Optimization
INTERPLAY: An Intelligent Model for Predicting Performance Degradation due to Multi-cache Way-disabling
Keyword: calibration
Modular Conformal Calibration
A Geometric Method for Improved Uncertainty Estimation in Real-time