Article
Computer Science, Artificial Intelligence
Wentai Wu, Ligang He, Weiwei Lin, Yi Su, Yuhua Cui, Carsten Maple, Stephen Jarvis
Summary: This paper presents a prediction-driven, unsupervised anomaly detection scheme using a backbone model that combines decomposition and inference of time series data. It proposes a novel metric called Local Trend Inconsistency (LTI) and an efficient detection algorithm. The experimental results show that the scheme outperforms existing algorithms in terms of the commonly used metric, Area Under Curve (AUC), while maintaining high efficiency.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Jin Fan, Zhentao Liu, Huifeng Wu, Jia Wu, Zhanyu Si, Peng Hao, Tom H. Luan
Summary: Anomaly detection of multivariate time series data has received significant research attention in recent years due to its wide applicability in various domains. However, existing methods face challenges in handling streaming data and the lack of expert knowledge in the training dataset. To address these challenges, we propose LUAD, a lightweight unsupervised anomaly detection scheme that combines Temporal Convolutional Network (TCN) and Variational AutoEncoder (VAE) techniques. Experimental results show that LUAD outperforms the state-of-the-art methods in both accuracy and efficiency.
Article
Computer Science, Theory & Methods
Xu Liu, Weiyou Liu, Xiaoqiang Di, Jinqing Li, Binbin Cai, Weiwu Ren, Huamin Yang
Summary: This paper proposes a network anomaly detection scheme LogNADS, which solves the problems in semantics-aware anomaly detection based on log by designing a novel log semantics representation method and an adaptive sequence data construction method. Experimental results demonstrate the effectiveness of LogNADS in detection accuracy and time cost, showing significant advantages over other state-of-the-art methods.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2021)
Article
Computer Science, Information Systems
Volodymyr Tkach, Anton Kudin, Victor R. R. Kebande, Oleksii Baranovskyi, Ivan Kudin
Summary: Anomaly detection in critical infrastructures is crucial for detecting threats and providing early warnings of potential cyber-attacks or infrastructure failures. This paper highlights the importance of Non-Pattern Anomaly Detection (NP-AD) in Time-Series and presents experimental results supporting its effectiveness in predicting future anomalies.
Article
Multidisciplinary Sciences
Gen Li, Jason J. Jung
Summary: The study proposed a dynamic graph embedding model, DynGPE, based on graph proximity for detecting abnormal climatic events in meteorological time series. Experimental results demonstrated that using DynGPE with three outlier detection methods can achieve better results than the baseline, with isolation forest providing the best performance.
Article
Computer Science, Information Systems
M. Van Onsem, D. De Paepe, S. Vanden Hautte, P. Bonte, V Ledoux, A. Lejon, F. Ongenae, D. Dreesen, S. Van Hoecke
Summary: This paper proposes a lightweight anomaly detection method that can observe crucial device metrics in real time for businesses to prevent downtime and data loss. The method is suitable for highly monitored production environments and does not require extensive historical data, while maintaining high accuracy.
COMPUTER COMMUNICATIONS
(2022)
Article
Computer Science, Information Systems
Jianming Lv, Yaquan Wang, Shengjing Chen
Summary: Existing multivariate time-series anomaly detection methods fail to address the temporal covariate shift problem, resulting in suboptimal detection performance. This paper proposes the DATECT framework, which combines a dilated convolution based AutoEncoder for robust anomaly measurement and an Adaptive Window Normalization method to improve the generalization capability. Additionally, Non-parametric Scan Statistics are utilized to reduce the side-effects of domain-specific dynamic noise. Experimental results demonstrate that DATECT significantly alleviates the performance drop caused by temporal covariate shift, outperforming the baseline in terms of detection performance and generalization, with an average improvement of 8.66% in F1-score and 1.18% in F1*-score (upper bound).
INFORMATION PROCESSING & MANAGEMENT
(2023)
Article
Computer Science, Artificial Intelligence
Zheng Liang, Hongzhi Wang, Xiaoou Ding, Tianyu Mu
Summary: The explosive growth of time series captured by sensors in industrial pipelines has led to the flourishing of intelligent industry. We propose a constraint hypergraph-based method for anomaly detection, developing strategies for adaptive determinative anomaly detection and anomaly pattern mining. Our approach effectively and efficiently works under different circumstances, as demonstrated with real world datasets from a powerplant, chemical production pipeline and hydraulic system.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Engineering, Multidisciplinary
Kai Yang, Shaoyu Dou, Pan Luo, Xin Wang, H. Vincent Poor
Summary: This paper introduces a sequence to Gaussian Mixture Model (seq2GMM) framework, aiming to identify anomalous and interesting time series within a network time series database. By developing a surrogate-based optimization algorithm, the model exhibits strong performance on multiple public benchmark datasets, outperforming state-of-the-art anomaly detection techniques.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Astha Garg, Wenyu Zhang, Jules Samaran, Ramasamy Savitha, Chuan-Sheng Foo
Summary: This article presents a systematic and comprehensive evaluation of unsupervised and semisupervised deep-learning-based methods for anomaly detection and diagnosis on multivariate time series data from cyberphysical systems. The study highlights that dynamic scoring functions work much better than static ones for multivariate time series anomaly detection, and the choice of scoring functions often matters more than the choice of the underlying model. Additionally, a new metric called the composite F-score (Fc_1) is proposed for evaluating time-series anomaly detection.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Multidisciplinary Sciences
Roozbeh Zarei, Guangyan Huang, Junfeng Wu
Summary: In this paper, a novel method called GraphTS is proposed for detecting subsequence anomalies in time series. The method uses a graph representation model and a 2D visualization technique to capture both recurrent and rare patterns in the time series. Experimental results show that the proposed method outperforms existing methods in terms of accuracy and efficiency in anomaly detection.
Article
Computer Science, Information Systems
Ane Blazquez-Garcia, Angel Conde, Usue Mori, Jose A. Lozano
Summary: The paper proposes a water leak detection method based on self-supervised classification of flow time series, which achieves the best balance between false positives and detected leaks in two water distribution networks when compared to two other methods in the literature.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Neeraj Chugh, Geetam Singh Tomar, Robin Singh Bhadoria, Neetesh Saxena
Summary: A new hybrid method called the data-driven zone-based routing protocol (DD-ZRP) is proposed for resource-constrained MANETs, incorporating anomaly detection schemes for security and energy awareness. The simulation results show improved detection ratio and performance for DD-ZRP compared to existing schemes, making it substantially better in terms of anomaly detection for security enhancement, energy efficiency, and optimization of available resources.
Article
Computer Science, Artificial Intelligence
Shenghua Liu, Bin Zhou, Quan Ding, Bryan Hooi, Zhengbo Zhang, Huawei Shen, Xueqi Cheng
Summary: Time series data naturally occur in various domains, but anomalous time series are rare. Existing approaches for anomaly detection are mostly based on supervised classification models, which require labeled anomalous data that are challenging to obtain in real-world scenarios. This paper proposes an unsupervised reconstruction model named BeatGAN, which learns to detect anomalies based on mostly normal data. BeatGAN utilizes adversarial learning to reconstruct and can work with both 1-d CNN and RNN. It identifies anomalies by detecting larger reconstruction errors and employs data augmentation with dynamic time warping for regularization and robustness. Experimental results demonstrate that BeatGAN achieves better accuracy and faster inference compared to other methods.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Emergency Medicine
Qian Cheng, Nilay Tanik Argon, Christopher Scott Evans, Yufeng Liu, Timothy F. Platts-Mills, Serhan Ziya
Summary: This study developed a novel predictive model for emergency department hourly occupancy using SARIMAX model with external regressors, providing up to 4-hour ahead predictions. By including external regressors of current ED occupancy, average department-wide ESI, and ED boarding total, the 24-SARIMAX model outperformed other forecasting methods in predicting ED occupancy.
AMERICAN JOURNAL OF EMERGENCY MEDICINE
(2021)
Article
Materials Science, Multidisciplinary
Abdelhakim Dorbane, Fouzi Harrou, Ying Sun
Summary: This study developed effective data-driven approaches using LSTM and GRU models to forecast the stress-strain curves of Al6061-T6 aluminum alloy base material and welded material under different temperature conditions. The results demonstrated the promising capacity of the GRU model in terms of efficiency and accuracy compared to the LSTM model.
JOURNAL OF MATERIALS ENGINEERING AND PERFORMANCE
(2023)
Article
Computer Science, Hardware & Architecture
K. Ramakrishna Kini, Muddu Madakyaru, Fouzi Harrou, Ying Sun
Summary: This article introduces a semisupervised machine learning approach that utilizes foot plantar pressure measurements to identify pediatric foot deformities. It combines PCA for feature extraction and the Kantorovich distance to effectively handle deformities.
IEEE DESIGN & TEST
(2023)
Article
Engineering, Electrical & Electronic
K. Ramakrishna Kini, Fouzi Harrou, Muddu Madakyaru, Farid Kadri, Ying Sun
Summary: Automatic and reliable detection of a person's sitting posture in a wheelchair is important for preventing health issues. This study proposes an unsupervised anomaly detection approach using pressure sensor data from the wheelchair and combines independent component analysis (ICA) with the Kantorovich Distance (KD) to improve anomaly detection. The ICA-KD approach shows promising performance in recognizing abnormal sitting posture.
IEEE INSTRUMENTATION & MEASUREMENT MAGAZINE
(2023)
Article
Biology
Zihao Wu, Carolina Euan, Rosa M. Crujeiras, Ying Sun
Summary: The directional wave spectrum (DWS) is a useful tool for marine studies and the design of maritime structures. This paper addresses the challenge of accounting for the circular nature of direction in the statistical estimation of DWS. It proposes using circular regression to smooth the observations and improve clustering algorithms for DWS analysis. Simulation studies and real data analysis demonstrate the effectiveness of the proposed circular smoother.
JOURNAL OF AGRICULTURAL BIOLOGICAL AND ENVIRONMENTAL STATISTICS
(2023)
Article
Engineering, Electrical & Electronic
Belkacem Khaldi, Fouzi Harrou, Abdelkader Dairi, Ying Sun
Summary: In this study, six powerful recurrent deep neural networks, including RNN, LSTM, GRU, ConvLSTM, Bidirectional LSTM (BiLSTM), and BiGRU, were explored and compared in forecasting the motion speed of miniature swarm mobile robots engaged in a simple aggregation formation task. The results showed the promising performance of DL for swarm motion forecasting, with BiGRU achieving the highest swarm motion speed forecasting performance in both no/with obstacles scenarios considered in this study.
JOURNAL OF ELECTRICAL ENGINEERING & TECHNOLOGY
(2023)
Article
Biochemistry & Molecular Biology
Slimane Laref, Fouzi Harrou, Bin Wang, Ying Sun, Amel Laref, Taous-Meriem Laleg-Kirati, Takashi Gojobori, Xin Gao
Summary: Favipiravir and Ebselen have been identified as potential antiviral drugs that can bind to a phosphorene nanocarrier. Machine learning models were used to train the Hamiltonian and interaction energy of the antiviral molecules, and Bayesian optimization was employed to improve prediction accuracy. Additionally, density functional theory calculations revealed the stability of the hybrid drug under different conditions.
Article
Medicine, General & Internal
Fouzi Harrou, Abdelkader Dairi, Abdelhakim Dorbane, Farid Kadri, Ying Sun
Summary: This study proposes a new method using KPCA and OCSVM to identify COVID-19 infections based on blood test data. The approach aims to differentiate healthy individuals from infected ones by detecting abnormal features using non-linear pattern analysis. The method is semi-supervised and achieves enhanced discrimination performance compared to other models. It achieved a high accuracy level with an AUC of 0.99 for distinguishing positive and negative samples based on the test results. This approach shows promise for detecting COVID-19 infections without labeled data.
Article
Energy & Fuels
K. Ramakrishna Kini, Fouzi Harrou, Muddu Madakyaru, Ying Sun
Summary: This paper presents a novel semi-supervised data-based monitoring technique for fault detection in wind turbines using SCADA data. The technique builds upon the Independent Component Analysis (ICA) approach to effectively capture non-Gaussian features for detecting different types of sensor faults. The fault detection process integrates fault indicators based on I2d, I2e, and squared prediction error (SPE), and combines them with a Double Exponential Weighted Moving Average (DEWMA) chart and kernel density estimation for establishing nonparametric thresholds.
Editorial Material
Energy & Fuels
Fouzi Harrou, Ying Sun, Bilal Taghezouit, Abdelkader Dairi
Article
Materials Science, Multidisciplinary
Souad Makhfi, Abdelhakim Dorbane, Fouzi Harrou, Ying Sun
Summary: Accurately predicting cutting forces in hard turning processes can lead to improved process control, reduced tool wear, and enhanced productivity. This study aims to predict machining force components during the hard turning of AISI 52100 bearing steel using machine learning models. The Gaussian process regression (GPR) and decision tree regression outperformed the other models, with averaged root-mean-square error values of 14.44 and 12.72, respectively. Feature selection was performed using two algorithms to identify the most important features impacting the cutting forces. This study's findings can be useful in optimizing cutting parameters for hard turning processes to select cutting forces, reduce tool wear, and minimize the generated heat during the machining process.
JOURNAL OF MATERIALS ENGINEERING AND PERFORMANCE
(2023)
Article
Green & Sustainable Science & Technology
Mohamed Zine, Fouzi Harrou, Mohammed Terbeche, Mohammed Bellahcene, Abdelkader Dairi, Ying Sun
Summary: This study presents an effective approach for assessing e-learning readiness by combining the ADKAR model and machine learning-based feature importance identification methods. The findings reveal that ability and knowledge are the most significant factors influencing students' e-learning readiness, suggesting that universities should focus on enhancing students' abilities and providing them with the necessary knowledge to increase their readiness for e-learning.
Article
Energy & Fuels
Fouzi Harrou, K. Ramakrishna Kini, Muddu Madakyaru, Ying Sun
Summary: Fault detection in wind turbines is crucial for their safety and optimal performance. Existing approaches face challenges in handling multivariate and non-Gaussian data, low sensitivity to small changes, and setting appropriate detection thresholds. This paper proposes a semi-supervised data-driven approach using supervisory control and data acquisition (SCADA) data, combining independent component analysis (ICA) and the Kantorovich Distance (KD)-based fault detection scheme. Experimental evaluations demonstrate the superior performance of the proposed approach.
SUSTAINABLE ENERGY GRIDS & NETWORKS
(2023)
Article
Computer Science, Artificial Intelligence
Rui Lv, Dingheng Wang, Jiangbin Zheng, Zhao-Xu Yang
Summary: In this paper, the authors investigate tensor decomposition for neural network compression. They analyze the convergence and precision of tensor mapping theory, validate the rationality of tensor mapping and its superiority over traditional tensor approximation based on the Lottery Ticket Hypothesis. They propose an efficient method called 3D-KCPNet to compress 3D convolutional neural networks using the Kronecker canonical polyadic (KCP) tensor decomposition. Experimental results show that 3D-KCPNet achieves higher accuracy compared to the original baseline model and the corresponding tensor approximation model.
Article
Computer Science, Artificial Intelligence
Xiangkun He, Zhongxu Hu, Haohan Yang, Chen Lv
Summary: In this paper, a novel constrained multi-objective reinforcement learning algorithm is proposed for personalized end-to-end robotic control with continuous actions. The approach trains a single model using constraint design and a comprehensive index to achieve optimal policies based on user-specified preferences.
Article
Computer Science, Artificial Intelligence
Zhijian Zhuo, Bilian Chen, Shenbao Yu, Langcai Cao
Summary: In this paper, a novel method called Expansion with Contraction Method for Overlapping Community Detection (ECOCD) is proposed, which utilizes non-negative matrix factorization to obtain disjoint communities and applies expansion and contraction processes to adjust the degree of overlap. ECOCD is applicable to various networks with different properties and achieves high-quality overlapping community detection.
Article
Computer Science, Artificial Intelligence
Yizhe Zhu, Chunhui Zhang, Jialin Gao, Xin Sun, Zihan Rui, Xi Zhou
Summary: In this work, the authors propose a Contrastive Spatio-Temporal Distilling (CSTD) approach to improve the detection of high-compressed deepfake videos. The approach leverages spatial-frequency cues and temporal-contrastive alignment to fully exploit spatiotemporal inconsistency information.
Review
Computer Science, Artificial Intelligence
Laijin Meng, Xinghao Jiang, Tanfeng Sun
Summary: This paper provides a review of coverless steganographic algorithms, including the development process, known contributions, and general issues in image and video algorithms. It also discusses the security of coverless steganography from theoretical analysis to actual investigation for the first time.
Article
Computer Science, Artificial Intelligence
Yajie Bao, Tianwei Xing, Xun Chen
Summary: Visual question answering requires processing multi-modal information and effective reasoning. Neural-symbolic learning is a promising method, but current approaches lack uncertainty handling and can only provide a single answer. To address this, we propose a confidence based neural-symbolic approach that evaluates NN inferences and conducts reasoning based on confidence.
Article
Computer Science, Artificial Intelligence
Anh H. Vo, Bao T. Nguyen
Summary: Interior style classification is an interesting problem with potential applications in both commercial and academic domains. This project proposes a method named ISC-DeIT, which combines data-efficient image transformer architectures and knowledge distillation, to address the interior style classification problem. Experimental results demonstrate a significant improvement in predictive accuracy compared to other state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Shashank Kotyan, Danilo Vasconcellos Vargas
Summary: This article introduces a novel augmentation technique called Dynamic Scanning Augmentation to improve the accuracy and robustness of Vision Transformer (ViT). The technique leverages dynamic input sequences to adaptively focus on different patches, resulting in significant changes in ViT's attention mechanism. Experimental results demonstrate that Dynamic Scanning Augmentation outperforms ViT in terms of both robustness to adversarial attacks and accuracy against natural images.
Article
Computer Science, Artificial Intelligence
Hiba Alqasir, Damien Muselet, Christophe Ducottet
Summary: The article proposes a solution to improve the learning process of a classification network by providing shape priors, reducing the need for annotated data. The solution is tested on cross-domain digit classification tasks and a video surveillance application.
Article
Computer Science, Artificial Intelligence
Dexiu Ma, Mei Liu, Mingsheng Shang
Summary: This paper proposes a method using neural dynamics solvers to solve infinity-norm optimization problems. Two improved solvers are constructed and their effectiveness and superiority are demonstrated through theoretical analysis and simulation experiments.
Article
Computer Science, Artificial Intelligence
Francesco Gregoretti, Giovanni Pezzulo, Domenico Maisto
Summary: Active Inference is a computational framework that uses probabilistic inference and variational free energy minimization to describe perception, planning, and action. cpp-AIF is a header-only C++ library that provides a powerful tool for implementing Active Inference for Partially Observable Markov Decision Processes through multi-core computing. It is cross-platform and improves performance, memory management, and usability compared to existing software.
Article
Computer Science, Artificial Intelligence
Zelin Ying, Dawei Cheng, Cen Chen, Xiang Li, Peng Zhu, Yifeng Luo, Yuqi Liang
Summary: This paper proposes a novel stock market trends prediction framework called SMART, which includes a self-supervised stock technical data sequence embedding model S3E. By training with multiple self-supervised auxiliary tasks, the model encodes stock technical data sequences into embeddings and uses the learned sequence embeddings for predicting stock market trends. Extensive experiments on China A-Shares market and NASDAQ market prove the high effectiveness of our model in stock market trends prediction, and its effectiveness is further validated in real-world applications in a leading financial service provider in China.
Article
Computer Science, Artificial Intelligence
Hao Li, Hao Jiang, Dongsheng Ye, Qiang Wang, Liang Du, Yuanyuan Zeng, Liu Yuan, Yingxue Wang, C. Chen
Summary: DHGAT1, a dynamic hyperbolic graph attention network, utilizes hyperbolic metric properties to embed dynamic graphs. It employs a spatiotemporal self-attention mechanism and weighted node representations, resulting in excellent performance in link prediction tasks.
Article
Computer Science, Artificial Intelligence
Jiehui Huang, Zhenchao Tang, Xuedong He, Jun Zhou, Defeng Zhou, Calvin Yu-Chian Chen
Summary: This study proposes a progressive learning multi-scale feature blending model for image deraining tasks. The model utilizes detail dilation and texture extraction to improve the restoration of rainy images. Experimental results show that the model achieves near state-of-the-art performance in rain removal tasks and exhibits better rain removal realism.
Article
Computer Science, Artificial Intelligence
Lizhi Liu, Zilin Gao, Yinhe Wang, Yongfu Li
Summary: This paper proposes a novel discrete-time interconnected model for depicting complex dynamical networks. The model consists of nodes and edges subsystems, which consider the dynamic characteristic of both nodes and edges. By designing control strategies and coupling modes, the stabilization and synchronization of the network are achieved. Simulation results demonstrate the effectiveness of the proposed methods.
