Discovery and Visualization of Nonstationary Causal Models
This work addresses the challenge of causal modeling in dynamic or heterogeneous environments, which is crucial for fields like time-series analysis and domain adaptation, though it appears incremental as it builds on existing constraint-based methods.
The paper tackles the problem of causal discovery in nonstationary data, where underlying generating processes change over time or across domains, by proposing a framework that detects nonstationary variables, determines causal directions, and visualizes causal modules, with experimental results demonstrating efficacy on synthetic and real-world datasets.
It is commonplace to encounter nonstationary data, of which the underlying generating process may change over time or across domains. The nonstationarity presents both challenges and opportunities for causal discovery. In this paper we propose a principled framework to handle nonstationarity, and develop some methods to address three important questions. First, we propose an enhanced constraint-based method to detect variables whose local mechanisms are nonstationary and recover the skeleton of the causal structure over observed variables. Second, we present a way to determine some causal directions by taking advantage of information carried by changing distributions. Third, we develop a method for visualizing the nonstationarity of causal modules. Experimental results on various synthetic and real-world data sets are presented to demonstrate the efficacy of our methods.