Orthogonality-Constrained Deep Instrumental Variable Model for Causal Effect Estimation

TL;DR

OC-DeepIV introduces an orthogonality-constrained neural network for causal effect estimation, improving accuracy and stability over existing methods by effectively modeling heterogeneity and reducing redundancy.

Contribution

The paper presents a novel neural network model with orthogonal constraints for causal inference, enhancing performance over prior deep learning approaches.

Findings

Outperforms DeepIV and DML in accuracy and stability

Effectively models heterogeneity with interaction features

Reduces redundancy through orthogonal regularization

Abstract

OC-DeepIV is a neural network model designed for estimating causal effects. It characterizes heterogeneity by adding interaction features and reduces redundancy through orthogonal constraints. The model includes two feature extractors, one for the instrumental variable Z and the other for the covariate X*. The training process is divided into two stages: the first stage uses the mean squared error (MSE) loss function, and the second stage incorporates orthogonal regularization. Experimental results show that this model outperforms DeepIV and DML in terms of accuracy and stability. Future research directions include applying the model to real-world problems and handling scenarios with multiple processing variables.