Coarse-to-Fine Learning of Dynamic Causal Structures
This work provides a more stable and efficient method for identifying fully dynamic causal structures, which is crucial for researchers and practitioners working with complex, time-varying systems where traditional stationary causality assumptions do not hold.
This paper addresses the challenge of learning dynamic causal structures in time series where causal relationships evolve over time. The proposed method, DyCausal, uses convolutional networks to identify causal patterns in coarse time windows and then refines these structures using linear interpolation to recover fine-grained, time-varying causal graphs. DyCausal outperforms existing methods on synthetic and real-world datasets, providing a stable and efficient solution for fully dynamic causal structure identification.
Learning the dynamic causal structure of time series is a challenging problem. Most existing approaches rely on distributional or structural invariance to uncover underlying causal dynamics, assuming stationary or partially stationary causality. However, these assumptions often conflict with the complex, time-varying causal relationships observed in real-world systems. This motivates the need for methods that address fully dynamic causality, where both instantaneous and lagged dependencies evolve over time. Such a setting poses significant challenges for the efficiency and stability of causal discovery. To address these challenges, we introduce DyCausal, a dynamic causal structure learning framework. DyCausal leverages convolutional networks to capture causal patterns within coarse-grained time windows, and then applies linear interpolation to refine causal structures at each time step, thereby recovering fine-grained and time-varying causal graphs. In addition, we propose an acyclic constraint based on matrix norm scaling, which improves efficiency while effectively constraining loops in evolving causal structures. Comprehensive evaluations on both synthetic and real-world datasets demonstrate that DyCausal achieves superior performance compared to existing methods, offering a stable and efficient approach for identifying fully dynamic causal structures from coarse to fine.