Transfer Learning for Individual Treatment Effect Estimation

TL;DR

This paper investigates transferring causal knowledge for individual treatment effect estimation, providing theoretical bounds and a practical framework that significantly reduces data needs through task similarity measures.

Contribution

It introduces a new theoretical analysis of ITE transfer feasibility and proposes the CITA measure for effective source task selection.

Findings

Up to 95% reduction in data required for ITE estimation.

Theoretical bounds support the feasibility of ITE transfer.

Empirical results validate the effectiveness of the proposed method.

Abstract

This work considers the problem of transferring causal knowledge between tasks for Individual Treatment Effect (ITE) estimation. To this end, we theoretically assess the feasibility of transferring ITE knowledge and present a practical framework for efficient transfer. A lower bound is introduced on the ITE error of the target task to demonstrate that ITE knowledge transfer is challenging due to the absence of counterfactual information. Nevertheless, we establish generalization upper bounds on the counterfactual loss and ITE error of the target task, demonstrating the feasibility of ITE knowledge transfer. Subsequently, we introduce a framework with a new Causal Inference Task Affinity (CITA) measure for ITE knowledge transfer. Specifically, we use CITA to find the closest source task to the target task and utilize it for ITE knowledge transfer. Empirical studies are provided, demonstrating the efficacy of the proposed method. We observe that ITE knowledge transfer can significantly (up to 95%) reduce the amount of data required for ITE estimation.