Robust model-free feature screening for ultrahigh dimensional surrogate data

This paper is concerned with the feature screening for the ultrahigh dimensional data with covariates missing at random, and some surrogate variables are available. We propose a marginal screening procedure based on the augmented inverse probability weighted methods and the nonparametric imputation technique. Our proposed screening method utilizes the surrogate information efficiently to overcome the missing data problem. It is model free and possesses the sure screening property under some regular conditions. Monte Carlo simulation studies and a real data application are conducted to examine the performance of the proposed procedure.