Feature screening and FDR control with knockoff features for ultrahigh-dimensional right-censored data

Abstract

A model-free feature screening method for ultrahigh-dimensional right-censored data is advocated. A two-step approach, with the help of knockoff features, is proposed to specify the threshold for feature screening such that the false discovery rate (FDR) is controlled under a prespecified level. The proposed two-step approach enjoys both a sure screening property with high probability and FDR control simultaneously if the prespecified FDR level is greater than or equal to 1/s, where s is the number of active features. The finite sample properties of the newly suggested method are assessed through simulation studies. An application to the mantle cell lymphoma (MCL) study demonstrates the utility of the proposed method in practice.