Cycle-Balanced Representation Learning For Counterfactual Inference

TL;DR

This paper introduces CBRE, a novel cycle-balanced representation learning framework that improves counterfactual inference from observational data by addressing distribution discrepancies and preserving data properties.

Contribution

It proposes a cycle-balanced adversarial learning framework that enhances counterfactual effect estimation by reducing information loss and aligning treatment and control group distributions.

Findings

CBRE outperforms state-of-the-art methods on real-world datasets.

The framework effectively reduces distribution discrepancy.

It preserves original data properties through cyclic information loops.

Abstract

With the widespread accumulation of observational data, researchers obtain a new direction to learn counterfactual effects in many domains (e.g., health care and computational advertising) without Randomized Controlled Trials(RCTs). However, observational data suffer from inherent missing counterfactual outcomes, and distribution discrepancy between treatment and control groups due to behaviour preference. Motivated by recent advances of representation learning in the field of domain adaptation, we propose a novel framework based on Cycle-Balanced REpresentation learning for counterfactual inference (CBRE), to solve above problems. Specifically, we realize a robust balanced representation for different groups using adversarial training, and meanwhile construct an information loop, such that preserve original data properties cyclically, which reduces information loss when transforming data into latent representation space.Experimental results on three real-world datasets demonstrate that CBRE matches/outperforms the state-of-the-art methods, and it has a great potential to be applied to counterfactual inference.