Coordinated Double Machine Learning
This is an incremental improvement for researchers in causal inference and machine learning, addressing bias in observational data analysis.
The paper tackles the problem of bias in treatment effect estimation using double machine learning by proposing a coordinated learning algorithm for deep neural networks, demonstrating improved empirical performance on simulated and real data.
Double machine learning is a statistical method for leveraging complex black-box models to construct approximately unbiased treatment effect estimates given observational data with high-dimensional covariates, under the assumption of a partially linear model. The idea is to first fit on a subset of the samples two non-linear predictive models, one for the continuous outcome of interest and one for the observed treatment, and then to estimate a linear coefficient for the treatment using the remaining samples through a simple orthogonalized regression. While this methodology is flexible and can accommodate arbitrary predictive models, typically trained independently of one another, this paper argues that a carefully coordinated learning algorithm for deep neural networks may reduce the estimation bias. The improved empirical performance of the proposed method is demonstrated through numerical experiments on both simulated and real data.