Multivariate Time Series Anomaly Detection via Dual-Branch Reconstruction and Autoregressive Flow-based Residual Density Estimation

TL;DR

This paper introduces DBR-AF, a novel multivariate time series anomaly detection framework that combines dual-branch reconstruction and autoregressive flow-based residual density estimation to improve accuracy and robustness.

Contribution

The paper presents a new framework that decouples variable correlation learning from statistical modeling and employs density estimation for better anomaly detection.

Findings

Achieves state-of-the-art results on seven benchmark datasets.

Validates the importance of core components through ablation studies.

Effectively distinguishes hard-to-reconstruct samples from true anomalies.

Abstract

Multivariate Time Series Anomaly Detection (MTSAD) is critical for real-world monitoring scenarios such as industrial control and aerospace systems. Mainstream reconstruction-based anomaly detection methods suffer from two key limitations: first, overfitting to spurious correlations induced by an overemphasis on cross-variable modeling; second, the generation of misleading anomaly scores by simply summing up multivariable reconstruction errors, which makes it difficult to distinguish between hard-to-reconstruct samples and genuine anomalies. To address these issues, we propose DBR-AF, a novel framework that integrates a dual-branch reconstruction (DBR) encoder and an autoregressive flow (AF) module. The DBR encoder decouples cross-variable correlation learning and intra-variable statistical property modeling to mitigate spurious correlations, while the AF module employs multiple stacked reversible transformations to model the complex multivariate residual distribution and further leverages density estimation to accurately identify normal samples with large reconstruction errors. Extensive experiments on seven benchmark datasets demonstrate that DBR-AF achieves state-of-the-art performance, with ablation studies validating the indispensability of its core components.