Causally-Constrained Probabilistic Forecasting for Time-Series Anomaly Detection

TL;DR

This paper presents a causally-constrained probabilistic forecasting framework using a Causally Guided Transformer for improved multivariate time-series anomaly detection, emphasizing causal interpretability and root-cause localization.

Contribution

It introduces a novel transformer-based model that integrates explicit causal graphs and uncertainty modeling for more robust and interpretable anomaly detection.

Findings

Achieved state-of-the-art F1-scores of 96.19% on ASD and 95.32% on SMD benchmarks.

Enhanced variable-level attribution quality in anomaly detection.

Demonstrated improved robustness and interpretability through causal priors.

Abstract

Anomaly detection in multivariate time series is a central challenge in industrial monitoring, as failures frequently arise from complex temporal dynamics and cross-sensor interactions. While recent deep learning models, including graph neural networks and Transformers, have demonstrated strong empirical performance, most approaches remain primarily correlational and offer limited support for causal interpretation and root-cause localization. This study introduces a causally-constrained probabilistic forecasting framework which is a Causally Guided Transformer (CGT) model for multivariate time-series anomaly detection, integrating an explicit time-lagged causal graph prior with deep sequence modeling. For each target variable, a dedicated forecasting block employs a hard parent mask derived from causal discovery to restrict the main prediction pathway to graph-supported causes, while a latent Gaussian head captures predictive uncertainty. To leverage residual correlational information without compromising the causal representation, a shadow auxiliary path with stop-gradient isolation and a safety-gated blending mechanism is incorporated to suppress non-causal contributions when reliability is low. Anomalies are identified using negative log-likelihood scores with adaptive streaming thresholding, and root-cause variables are determined through per-dimension probabilistic attribution and counterfactual clamping. Experiments on the ASD and SMD benchmarks indicate that the proposed method achieves state-of-the-art detection performance, with F1-scores of 96.19% on ASD and 95.32% on SMD, and enhances variable-level attribution quality. These findings suggest that causal structural priors can improve both robustness and interpretability in detecting deep anomalies in multivariate sensor systems.