Correlation based LSTM for Multivariate Time Series Anomaly Detection

With the proliferation of multivariate time series data in various fields, this paper proposes an anomaly detection method based on modelling the correlation of multivariate time series. The correlation between multivariate time series is quantified as a correlation matrix, which is combined with th...

Full description

Saved in:
Bibliographic Details
Published in2024 5th International Conference on Computer Vision, Image and Deep Learning (CVIDL) pp. 380 - 383
Main Authors Zheng, Jiuhao, Duan, Jiangyong, Zhang, Ke, Ma, Zhongsong
Format Conference Proceeding
LanguageEnglish
Published IEEE 19.04.2024
Subjects
Anomaly detection
Correlation
Deep learning
Feature extraction
Multivariate time series
Noise
Predictive models
Time series analysis
Time series correlation
Training
Online AccessGet full text

Cover

More Information
Summary:With the proliferation of multivariate time series data in various fields, this paper proposes an anomaly detection method based on modelling the correlation of multivariate time series. The correlation between multivariate time series is quantified as a correlation matrix, which is combined with the original features as enhanced features for training anomaly detection models. Meanwhile, a deep learning network model is utilized to capture the temporal dependencies within sequences, thereby detecting anomalies caused by changes in inter-channel correlations. In experiments, our method produces higher precision, recall, and F1 scores compared to other state-of-the-art models, validating the effectiveness of the proposed method.
DOI:10.1109/CVIDL62147.2024.10603763