Model-assisted inference for treatment effects using regularized calibrated estimation with high-dimensional data

Zhiqiang Tan

Consider the problem of estimating average treatment effects when a large number of covariates are used to adjust for possible confounding through outcome regression and propensity…