Causal Temporal Regime Structure Learning

TL;DR

This paper introduces CASTOR, a novel framework for learning causal structures in multivariate time series with multiple regimes, effectively identifying regime boundaries and causal graphs even with changing dynamics.

Contribution

The work presents a new method that jointly learns regime segmentation and causal graphs, with proven identifiability and superior performance over existing models.

Findings

CASTOR outperforms existing models in regime detection and causal graph learning.

It effectively handles both linear and nonlinear causal relationships.

The method is validated on synthetic and real-world datasets.

Abstract

Understanding causal relationships in multivariate time series is essential for predicting and controlling dynamic systems in fields like economics, neuroscience, and climate science. However, existing causal discovery methods often assume stationarity, limiting their effectiveness when time series consist of sequential regimes, consecutive temporal segments with unknown boundaries and changing causal structures. In this work, we firstly introduce a framework to describe and model such time series. Then, we present CASTOR, a novel method that concurrently learns the Directed Acyclic Graph (DAG) for each regime while determining the number of regimes and their sequential arrangement. CASTOR optimizes the data log-likelihood using an expectation-maximization algorithm, alternating between assigning regime indices (expectation step) and inferring causal relationships in each regime (maximization step). We establish the identifiability of the regimes and DAGs within our framework. Extensive experiments show that CASTOR consistently outperforms existing causal discovery models in detecting different regimes and learning their DAGs across various settings, including linear and nonlinear causal relationships, on both synthetic and real world datasets.