Causal Discovery with Heterogeneous Observational Data

TL;DR

This paper introduces a Bayesian method for causal discovery from heterogeneous observational data, leveraging data variability to identify causal structures, including cyclic ones, with uncertainty quantification.

Contribution

It proposes a novel approach that models causal effects as functions of exogenous covariates, enabling structure learning from heterogeneous data, unlike traditional methods assuming homogeneity.

Findings

Effective in identifying cyclic causal structures

Provides uncertainty quantification in causal inference

Validated through simulations and real-world data

Abstract

We consider the problem of causal discovery (structure learning) from heterogeneous observational data. Most existing methods assume a homogeneous sampling scheme, which leads to misleading conclusions when violated in many applications. To this end, we propose a novel approach that exploits data heterogeneity to infer possibly cyclic causal structures from causally insufficient systems. The core idea is to model the direct causal effects as functions of exogenous covariates that properly explain data heterogeneity. We investigate structure identifiability properties of the proposed model. Structure learning is carried out in a fully Bayesian fashion, which provides natural uncertainty quantification. We demonstrate its utility through extensive simulations and a real-world application.