Counterfactual Generative Modeling with Variational Causal Inference

TL;DR

This paper introduces a novel variational Bayesian framework for counterfactual generative modeling that effectively predicts individual outcomes under interventions, especially with high-dimensional data and limited covariates.

Contribution

It proposes a new theoretical framework that enables end-to-end counterfactual supervision without counterfactual samples and promotes disentangled noise for causal effect identification.

Findings

Outperforms state-of-the-art models on multiple benchmarks.

Effectively handles high-dimensional outcomes with limited covariates.

Enables counterfactual prediction without requiring counterfactual data.

Abstract

Estimating an individual's counterfactual outcomes under interventions is a challenging task for traditional causal inference and supervised learning approaches when the outcome is high-dimensional (e.g. gene expressions, facial images) and covariates are relatively limited. In this case, to predict one's outcomes under counterfactual treatments, it is crucial to leverage individual information contained in the observed outcome in addition to the covariates. Prior works using variational inference in counterfactual generative modeling have been focusing on neural adaptations and model variants within the conditional variational autoencoder formulation, which we argue is fundamentally ill-suited to the notion of counterfactual in causal inference. In this work, we present a novel variational Bayesian causal inference framework and its theoretical backings to properly handle counterfactual generative modeling tasks, through which we are able to conduct counterfactual supervision end-to-end during training without any counterfactual samples, and encourage disentangled exogenous noise abduction that aids the correct identification of causal effect in counterfactual generations. In experiments, we demonstrate the advantage of our framework compared to state-of-the-art models in counterfactual generative modeling on multiple benchmarks.