Representation learning with CGAN for casual inference

TL;DR

This paper introduces a novel method using CGAN-based adversarial representation learning to improve causal inference, supported by theoretical analysis showing its effectiveness when distributions are balanced.

Contribution

It proposes a new adversarial representation learning approach with CGAN for causal inference, backed by theoretical demonstration of its feasibility.

Findings

Theoretically demonstrates the feasibility of finding suitable representations with CGAN.

Shows that balanced distributions enable the ideal representation function.

Provides a foundation for further research in causal inference using CGAN.

Abstract

Conditional Generative Adversarial Nets (CGAN) is often used to improve conditional image generation performance. However, there is little research on Representation learning with CGAN for causal inference. This paper proposes a new method for finding representation learning functions by adopting the adversarial idea. We apply the pattern of CGAN and theoretically emonstrate the feasibility of finding a suitable representation function in the context of two distributions being balanced. The theoretical result shows that when two distributions are balanced, the ideal representation function can be found and thus can be used to further research.