Controlling for discrete unmeasured confounding in nonlinear causal models

TL;DR

This paper introduces a novel method to detect and correct for unmeasured discrete confounding in nonlinear causal models, leveraging deep latent variable models and flow-based algorithms, with promising results on synthetic and real data.

Contribution

It extends identifiability results to accommodate unmeasured discrete confounding in nonlinear causal models using a flow-based estimation approach.

Findings

Effective deconfounding demonstrated on synthetic data

Successful application to real-world datasets

Theoretical detection of confounding under specific assumptions

Abstract

Unmeasured confounding is a major challenge for identifying causal relationships from non-experimental data. Here, we propose a method that can accommodate unmeasured discrete confounding. Extending recent identifiability results in deep latent variable models, we show theoretically that confounding can be detected and corrected under the assumption that the observed data is a piecewise affine transformation of a latent Gaussian mixture model and that the identity of the mixture components is confounded. We provide a flow-based algorithm to estimate this model and perform deconfounding. Experimental results on synthetic and real-world data provide support for the effectiveness of our approach.