StackVAE-G: An efficient and interpretable model for time series anomaly detection

被引:4
|
作者
Li, Wenkai [1 ,4 ]
Hu, Wenbo [2 ]
Chen, Ting [1 ]
Chen, Ning [1 ]
Feng, Cheng [3 ,4 ]
机构
[1] Tsinghua Univ, Tsinghua Bosch Joint ML Ctr, High Performance Computing Ctr,BNRist Ctr, Dept Comp Sci & Tech,Inst AI,THBI Lab, Beijing, Peoples R China
[2] Hefei Univ Technol, Sch Comp & Informat, Hefei, Peoples R China
[3] Siemens AG, Beijing, Peoples R China
[4] THU Siemens Joint Res Ctr Ind Intelligence & Inter, Beijing, Peoples R China
来源
AI OPEN | 2022年 / 3卷
关键词
Time-series; Anomaly detection; Autoencoders; Graph neural network;
D O I
10.1016/j.aiopen.2022.07.001
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Recent studies have shown that autoencoder-based models can achieve superior performance on anomaly detection tasks due to their excellent ability to fit complex data in an unsupervised manner. In this work, we propose a novel autoencoder-based model, named StackVAE-G that can significantly bring the efficiency and interpretability to multivariate time series anomaly detection. Specifically, we utilize the similarities across the time series channels by the stacking block -wise reconstruction with a weight -sharing scheme to reduce the size of learned models and also relieve the overfitting to unknown noises in the training data. We also leverage a graph learning module to learn a sparse adjacency matrix to explicitly capture the stable interrelation structure among multiple time series channels for the interpretable pattern reconstruction of interrelated channels. Combining these two modules, we introduce the stacking block -wise VAE (variational autoencoder) with GNN (graph neural network) model for multivariate time series anomaly detection. We conduct extensive experiments on three commonly used public datasets, showing that our model achieves comparable (even better) performance with the state-of-the-art models and meanwhile requires much less computation and memory cost. Furthermore, we demonstrate that the adjacency matrix learned by our model accurately captures the interrelation among multiple channels, and can provide valuable information for failure diagnosis applications.
引用
收藏
页码:101 / 110
页数:10
相关论文
共 50 条
  • [1] An anomaly detection model for multivariate time series with anomaly perception
    Wei, Dong
    Sun, Wu
    Zou, Xiaofeng
    Ma, Dan
    Xu, Huarong
    Chen, Panfeng
    Yang, Chaoshu
    Chen, Mei
    Li, Hui
    [J]. PEERJ COMPUTER SCIENCE, 2024, 10
  • [2] An anomaly detection model for multivariate time series with anomaly perception
    Wei, Dong
    Sun, Wu
    Zou, Xiaofeng
    Ma, Dan
    Xu, Huarong
    Chen, Panfeng
    Yang, Chaoshu
    Chen, Mei
    Li, Hui
    [J]. PeerJ Computer Science, 2024, 10
  • [3] EAD: An Efficient Anomaly Detection Algorithm for Multivariate Time Series
    Ma, Dehong
    Ding, Bo
    Feng, Dawei
    Liu, Hui
    [J]. 2021 IEEE 33RD INTERNATIONAL CONFERENCE ON TOOLS WITH ARTIFICIAL INTELLIGENCE (ICTAI 2021), 2021, : 609 - 613
  • [4] Data-Efficient and Interpretable Tabular Anomaly Detection
    Chang, Chun-Hao
    Yoon, Jinsung
    Arik, Sercan O.
    Udell, Madeleine
    Pfister, Tomas
    [J]. PROCEEDINGS OF THE 29TH ACM SIGKDD CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, KDD 2023, 2023, : 190 - 201
  • [5] Clustering model for anomaly detection in time series sets
    Leng, Mingwei
    Lai, Xinsheng
    Liu, Yingdong
    [J]. Journal of Information and Computational Science, 2010, 7 (01): : 85 - 93
  • [6] A Generative Model for Anomaly Detection in Time Series Data
    Hoh, Maximilian
    Schoettl, Alfred
    Schaub, Henry
    Wenninger, Franz
    [J]. 3RD INTERNATIONAL CONFERENCE ON INDUSTRY 4.0 AND SMART MANUFACTURING, 2022, 200 : 629 - 637
  • [7] GRU-Based Interpretable Multivariate Time Series Anomaly Detection in Industrial Control System
    Tang, Chaofan
    Xu, Lijuan
    Yang, Bo
    Tang, Yongwei
    Zhao, Dawei
    [J]. COMPUTERS & SECURITY, 2023, 127
  • [8] Towards Efficient Deep Autoencoders for Multivariate Time Series Anomaly Detection
    Pietron, Marcin
    Zurek, Dominik
    Faber, Kamil
    Corizzo, Roberto
    [J]. COMPUTATIONAL SCIENCE, ICCS 2024, PT III, 2024, 14834 : 461 - 469
  • [9] ReFuzzTiD: A Recurrent Neurofuzzy Model for Anomaly Detection in Time Series
    Kandilogiannakis, George
    Mastorocostas, Paris
    [J]. 2020 INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS (IJCNN), 2020,
  • [10] Model-centered Ensemble for Anomaly Detection in Time Series
    Trentini, Erick L.
    Coelho da Silva, Ticiana L.
    Melo Junior, Leopoldo
    de Macedo, Jose F.
    [J]. ICAART: PROCEEDINGS OF THE 12TH INTERNATIONAL CONFERENCE ON AGENTS AND ARTIFICIAL INTELLIGENCE, VOL 2, 2020, : 700 - 707