Estimating Average Treatment Effects via Orthogonal Regularization

TL;DR

This paper introduces a novel regularization framework that exploits the unconfoundedness assumption by formalizing it as an orthogonality constraint, leading to improved estimation of treatment effects from observational data.

Contribution

It proposes a new regularization method based on orthogonality constraints and develops deep orthogonal networks (DONUT) for better treatment effect estimation.

Findings

DONUT outperforms existing methods on benchmark datasets.

Orthogonality constraints improve outcome estimation accuracy.

Deep orthogonal networks effectively leverage unconfoundedness.

Abstract

Decision-making often requires accurate estimation of treatment effects from observational data. This is challenging as outcomes of alternative decisions are not observed and have to be estimated. Previous methods estimate outcomes based on unconfoundedness but neglect any constraints that unconfoundedness imposes on the outcomes. In this paper, we propose a novel regularization framework for estimating average treatment effects that exploits unconfoundedness. To this end, we formalize unconfoundedness as an orthogonality constraint, which ensures that the outcomes are orthogonal to the treatment assignment. This orthogonality constraint is then included in the loss function via a regularization. Based on our regularization framework, we develop deep orthogonal networks for unconfounded treatments (DONUT), which learn outcomes that are orthogonal to the treatment assignment. Using a variety of benchmark datasets for estimating average treatment effects, we demonstrate that DONUT outperforms the state-of-the-art substantially.