Wireless Sensor Networks anomaly detection using Machine Learning: A Survey
:brain: Authors: Ahsnaul Haque, Md Naseef-Ur-Rahman Chowdhury, Hamdy Soliman, Mohammad Sahinur Hossen, Tanjim Fatima, Imtiaz Ahmed :paw_prints: Details:arXiv:2303.08823 :ramen: tl;dr: Wireless Sensor Networks (WSNs) have become increasingly valuable in various civil/military applications like industrial process control, civil engineering applications such as buildings structural strength monitoring, environmental monitoring, border intrusion, IoT (Internet of Things), and healthcare. However, the sensed data generated by WSNs is often noisy and unreliable, making it a challenge to detect and diagnose anomalies. Machine learning (ML) techniques have been widely used to address this problem by detecting and identifying unusual patterns in the detected data. This survey paper provides an overview of the state of the art applications of ML techniques for data anomaly detection in WSN domains. We first introduce the characteristics of WSN-based ML techniques such as supervised, unsupervised, and semi-supervised learning that have been applied to WSN sensors. We also compare different ML-based approaches and their performance evaluation metrics. Finally, we discuss open research challenges and future directions for applying ML techniques in WNSs sensed data anomalies. In addition, our paper provides a detailed description of the current state of ML and the challenges of ML.... :sunflower: Comments: 19 pages, 4 figures, IntelliSys 2023 :seedling: Read more →
CoLo-CAM: Class Activation Mapping for Object Co-Localization in Weakly-Labeled Unconstrained Videos
:brain: Authors: Soufiane Belharbi, Shakeeb Murtaza, Marco Pedersoli, Ismail Ben Ayed, Luke McCaffrey, Eric Granger :paw_prints: Details:arXiv:2303.09044 :ramen: tl;dr: Weakly-supervised video object localization (WSVOL) methods often rely on visual and motion cues only, making them susceptible to inaccurate localization. Recently, discriminative models via a temporal class activation mapping (CAM) method have been explored. Although results are promising, objects are assumed to have minimal movement leading to degradation in performance for relatively long-term dependencies. In this paper, a novel CoLo-CAM method for object localization is proposed to leverage spatiotemporal information in activation maps without any assumptions about object movement. Over a given sequence of frames, explicit joint learning of localization is produced across these maps based on color cues, by assuming an object has similar color across frames. The CAMs' activations are constrained to activate similarly over pixels with similar colors, achieving co-localization. This joint learning creates direct communication among pixels across all image locations, and over all frames, allowing for transfer, aggregation, and correction of learned localization. This is achieved by minimizing a color term of a CRF loss over joint images/maps. In addition to our multi-frame constraint, we impose per-frame local constraints including pseudo-labels, and CRF gain in combination with a global size constraint to improve... :sunflower: Comments: 16 pages, 8 figures :seedling: Read more →
Web and Mobile Platforms for Managing Elections based on IoT And Machine Learning Algorithms
:brain: Authors: G. M. I. K. Galagoda, W. M. C. A. Karunarathne, R. S. Bates, K. M. H. V. P. Gangathilaka, Kanishka Yapa, Erandika Gamage :paw_prints: Details:arXiv:2303.09045 :ramen: tl;dr: The global pandemic situation has severely affected all countries. As a result, almost all countries had to adjust to online technologies to continue their processes. In addition, Sri Lanka is yearly spending ten billion on elections. We have examined a proper way of minimizing the cost of hosting these events online. To solve the existing problems and increase the time and cost reduction we have used IoT and ML-based technologies. IoT-based data will identify, register, and be used to secure from fraud, while ML algorithms manipulate the election data and produce winning predictions, weather-based voters attendance, and election violence. All the data will be saved in cloud computing and a standard database to store and access the data. This study mainly focuses on four aspects of an E-voting system. The most frequent problems across the world in E-Voting are the security, accuracy, and reliability of the systems. E-government systems must be secured against various cyber-attacks and ensure that only authorized users can access valuable, and sometimes sensitive information. Being able to access a system without passwords but using biometric details has been there for a while now, however, our proposed system has a different approach to taking the credentials, processing, and combining the images,... :seedling: Read more →
VFP: Converting Tabular Data for IIoT into Images Considering Correlations of Attributes for Convolutional Neural Networks
:brain: Authors: Jong-Ik Park, Cheol-Ho Hong :paw_prints: Details:arXiv:2303.09068 :ramen: tl;dr: For tabular data generated from IIoT devices, traditional machine learning (ML) techniques based on the decision tree algorithm have been employed. However, these methods have limitations in processing tabulardata where real number attributes dominate. To address this issue, DeepInsight, REFINED, and IGTD were proposed to convert tabularData into images for utilizing convolutional neural networks (CNNs). They gather similar features in some specific spots of an image to make the converted image look like an actual image. Gathering similar features contrasts with traditional ML techniques for tabular Data, which drops some highly correlated attributes to avoid overfitting. Also, previous converting methods fixed the image size, and there are wasted or insufficient pixels according to the number of attributes. Therefore, this paper proposes a new converting method, Vortex Feature Positioning (VFP). VFP considers the correlation of features and places similar features far away from each. Features are positioned in the vortex shape from the center of the image, and the number is used to determine the size of a given image. VFP shows better test performance than traditional ML technique for Tabular Data. The paper compares the results of VFP in five datasets Iris, Wine, Dry Bean... :seedling: Read more →
danielphan2003
commented
1 year ago
Keyword: iot (4 papers)
Wireless Sensor Networks anomaly detection using Machine Learning: A Survey
:brain: Authors: Ahsnaul Haque, Md Naseef-Ur-Rahman Chowdhury, Hamdy Soliman, Mohammad Sahinur Hossen, Tanjim Fatima, Imtiaz Ahmed:paw_prints: Details: arXiv:2303.08823
:ramen: tl;dr: Wireless Sensor Networks (WSNs) have become increasingly valuable in various civil/military applications like industrial process control, civil engineering applications such as buildings structural strength monitoring, environmental monitoring, border intrusion, IoT (Internet of Things), and healthcare. However, the sensed data generated by WSNs is often noisy and unreliable, making it a challenge to detect and diagnose anomalies. Machine learning (ML) techniques have been widely used to address this problem by detecting and identifying unusual patterns in the detected data. This survey paper provides an overview of the state of the art applications of ML techniques for data anomaly detection in WSN domains. We first introduce the characteristics of WSN-based ML techniques such as supervised, unsupervised, and semi-supervised learning that have been applied to WSN sensors. We also compare different ML-based approaches and their performance evaluation metrics. Finally, we discuss open research challenges and future directions for applying ML techniques in WNSs sensed data anomalies. In addition, our paper provides a detailed description of the current state of ML and the challenges of ML....
:sunflower: Comments: 19 pages, 4 figures, IntelliSys 2023
:seedling: Read more →
CoLo-CAM: Class Activation Mapping for Object Co-Localization in Weakly-Labeled Unconstrained Videos
:brain: Authors: Soufiane Belharbi, Shakeeb Murtaza, Marco Pedersoli, Ismail Ben Ayed, Luke McCaffrey, Eric Granger:paw_prints: Details: arXiv:2303.09044
:ramen: tl;dr: Weakly-supervised video object localization (WSVOL) methods often rely on visual and motion cues only, making them susceptible to inaccurate localization. Recently, discriminative models via a temporal class activation mapping (CAM) method have been explored. Although results are promising, objects are assumed to have minimal movement leading to degradation in performance for relatively long-term dependencies. In this paper, a novel CoLo-CAM method for object localization is proposed to leverage spatiotemporal information in activation maps without any assumptions about object movement. Over a given sequence of frames, explicit joint learning of localization is produced across these maps based on color cues, by assuming an object has similar color across frames. The CAMs' activations are constrained to activate similarly over pixels with similar colors, achieving co-localization. This joint learning creates direct communication among pixels across all image locations, and over all frames, allowing for transfer, aggregation, and correction of learned localization. This is achieved by minimizing a color term of a CRF loss over joint images/maps. In addition to our multi-frame constraint, we impose per-frame local constraints including pseudo-labels, and CRF gain in combination with a global size constraint to improve...
:sunflower: Comments: 16 pages, 8 figures
:seedling: Read more →
Web and Mobile Platforms for Managing Elections based on IoT And Machine Learning Algorithms
:brain: Authors: G. M. I. K. Galagoda, W. M. C. A. Karunarathne, R. S. Bates, K. M. H. V. P. Gangathilaka, Kanishka Yapa, Erandika Gamage:paw_prints: Details: arXiv:2303.09045
:ramen: tl;dr: The global pandemic situation has severely affected all countries. As a result, almost all countries had to adjust to online technologies to continue their processes. In addition, Sri Lanka is yearly spending ten billion on elections. We have examined a proper way of minimizing the cost of hosting these events online. To solve the existing problems and increase the time and cost reduction we have used IoT and ML-based technologies. IoT-based data will identify, register, and be used to secure from fraud, while ML algorithms manipulate the election data and produce winning predictions, weather-based voters attendance, and election violence. All the data will be saved in cloud computing and a standard database to store and access the data. This study mainly focuses on four aspects of an E-voting system. The most frequent problems across the world in E-Voting are the security, accuracy, and reliability of the systems. E-government systems must be secured against various cyber-attacks and ensure that only authorized users can access valuable, and sometimes sensitive information. Being able to access a system without passwords but using biometric details has been there for a while now, however, our proposed system has a different approach to taking the credentials, processing, and combining the images,...
:seedling: Read more →
VFP: Converting Tabular Data for IIoT into Images Considering Correlations of Attributes for Convolutional Neural Networks
:brain: Authors: Jong-Ik Park, Cheol-Ho Hong:paw_prints: Details: arXiv:2303.09068
:ramen: tl;dr: For tabular data generated from IIoT devices, traditional machine learning (ML) techniques based on the decision tree algorithm have been employed. However, these methods have limitations in processing tabulardata where real number attributes dominate. To address this issue, DeepInsight, REFINED, and IGTD were proposed to convert tabularData into images for utilizing convolutional neural networks (CNNs). They gather similar features in some specific spots of an image to make the converted image look like an actual image. Gathering similar features contrasts with traditional ML techniques for tabular Data, which drops some highly correlated attributes to avoid overfitting. Also, previous converting methods fixed the image size, and there are wasted or insufficient pixels according to the number of attributes. Therefore, this paper proposes a new converting method, Vortex Feature Positioning (VFP). VFP considers the correlation of features and places similar features far away from each. Features are positioned in the vortex shape from the center of the image, and the number is used to determine the size of a given image. VFP shows better test performance than traditional ML technique for Tabular Data. The paper compares the results of VFP in five datasets Iris, Wine, Dry Bean...
:seedling: Read more →