Causal Disentanglement Learning for Accurate Anomaly Detection in Multivariate Time Series

TL;DR

CDRL4AD addresses anomaly detection in multivariate time series by learning causally disentangled representations that align latent factors with time-lagged causal mechanisms. It constructs a temporal heterogeneous graph to capture causal, correlation, and temporal dependencies, and integrates causal discovery with a multi-head VAE to produce interpretable latent factors. Joint prediction and reconstruction losses enable accurate anomaly scoring and root-cause analysis, with a GRU-based output layer aggregating diverse representations. Empirical results on eight real-world datasets show state-of-the-art detection performance and improved root-cause analysis, with competitive time complexity suitable for real-time deployment.

Abstract

Disentangling complex causal relationships is important for accurate detection of anomalies. In multivariate time series analysis, dynamic interactions among data variables over time complicate the interpretation of causal relationships. Traditional approaches assume statistical independence between variables in unsupervised settings, whereas recent methods capture feature correlations through graph representation learning. However, their representations fail to explicitly infer the causal relationships over different time periods. To solve the problem, we propose Causally Disentangled Representation Learning for Anomaly Detection (CDRL4AD) to detect anomalies and identify their causal relationships in multivariate time series. First, we design the causal process as model input, the temporal heterogeneous graph, and causal relationships. Second, our representation identifies causal relationships over different time periods and disentangles latent variables to infer the corresponding causal factors. Third, our experiments on real-world datasets demonstrate that CDRL4AD outperforms state-of-the-art methods in terms of accuracy and root cause analysis. Fourth, our model analysis validates hyperparameter sensitivity and the time complexity of CDRL4AD. Last, we conduct a case study to show how our approach assists human experts in diagnosing the root causes of anomalies.

Figures (4)

• Figure 1: The overview of our proposed framework, CDRL4AD
• Figure 2: Sensitivity Analysis
• Figure 3: Example of anomaly diagnosis process
• Figure 4: Example of causal discovery process