Self-Perturbed Anomaly-Aware Graph Dynamics for Multivariate Time-Series Anomaly Detection

See-Kiong Ng (National University of Singapore) · Jinyu Cai (National University of Singapore) · Yuan Xie (East Normal China University) · Glynnis Lim (National University of Singapore) · Yifang Yin (I2R, A*STAR) · Roger Zimmermann (National University of Singapore)

anomaly detectionauxiliary signalsbenchmark datasets class imbalanceclassifier trainingdynamic inter-variable correlationsgraph construction modulemultivariate time-seriesreconstruction residualsreconstruction-based methodsself-perturbation modulespatial convolutionsspatio-temporal anomaly detectionstatic graph-based methodstemporal convolutions

Detecting anomalies in multivariate time-series data is an essential task across various domains, yet there are unresolved challenges such as (1) severe class imbalance between normal and anomalous data due to rare anomaly availability in the real world; (2) limited adaptability of the static graph-based methods to dynamically changing inter-variable correlations; and (3) neglect of subtle anomalies due to overfitting to normal patterns in reconstruction-based methods. To tackle these issues, we propose Self-Perturbed Anomaly-Aware Graph Dynamics (SPAGD), a framework for time-series anomaly detection. SPAGD employs a self-perturbation module that generates self-perturbed time series from the reconstruction process of normal ones, which provide auxiliary signals to alleviate class imbalance during training. Concurrently, an anomaly-aware graph construction module is proposed to dynamically adjust the graph structure by leveraging the reconstruction residuals of self-perturbed time series, thereby emphasizing the inter-variable disruptions induced by anomalous candidates. A unified spatio-temporal anomaly detection module then integrates both spatial and temporal convolutions to train a classifier that distinguishes normal time series from the auxiliary self-perturbed samples. Extensive experiments across multiple benchmark datasets demonstrate the effectiveness of SPAGD compared to state-of-the-art baselines.