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

arXiv cs.AI / 4/13/2026

💬 OpinionIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper targets multivariate time series anomaly detection (MTSAD) in monitoring-critical settings such as industrial control and aerospace systems.
It identifies two shortcomings in common reconstruction-based approaches: overfitting to spurious cross-variable correlations and unreliable anomaly scoring from naive sums of reconstruction errors.
DBR-AF introduces a dual-branch reconstruction (DBR) encoder that separates cross-variable correlation learning from intra-variable statistical property modeling to reduce spurious correlations.
It adds an autoregressive flow (AF) module with stacked reversible transformations to model the residual distribution and perform density estimation to better distinguish true anomalies from hard-to-reconstruct samples.
Experiments on seven benchmark datasets report state-of-the-art results, and ablation studies confirm the necessity of DBR and AF components.

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.