Estimating Categorical Counterfactuals via Deep Twin Networks

TL;DR

This paper introduces deep twin networks for estimating categorical counterfactuals, establishing a new principle called counterfactual ordering to ensure trustworthy inference in complex domains.

Contribution

It proposes the concept of counterfactual ordering for categorical variables and develops deep twin networks to perform counterfactual inference without abduction, action, and prediction steps.

Findings

Accurate estimation of counterfactual probabilities in real-world data

Demonstrates issues in counterfactual reasoning without counterfactual ordering

Validates approach across medicine, epidemiology, and finance datasets

Abstract

Counterfactual inference is a powerful tool, capable of solving challenging problems in high-profile sectors. To perform counterfactual inference, one requires knowledge of the underlying causal mechanisms. However, causal mechanisms cannot be uniquely determined from observations and interventions alone. This raises the question of how to choose the causal mechanisms so that resulting counterfactual inference is trustworthy in a given domain. This question has been addressed in causal models with binary variables, but the case of categorical variables remains unanswered. We address this challenge by introducing for causal models with categorical variables the notion of counterfactual ordering, a principle that posits desirable properties causal mechanisms should posses, and prove that it is equivalent to specific functional constraints on the causal mechanisms. To learn causal mechanisms satisfying these constraints, and perform counterfactual inference with them, we introduce deep twin networks. These are deep neural networks that, when trained, are capable of twin network counterfactual inference -- an alternative to the abduction, action, & prediction method. We empirically test our approach on diverse real-world and semi-synthetic data from medicine, epidemiology, and finance, reporting accurate estimation of counterfactual probabilities while demonstrating the issues that arise with counterfactual reasoning when counterfactual ordering is not enforced.