A0334
Title: CoxKnockoff: Controlled feature selection for the Cox model using knockoffs
Authors: Daoji Li - California State University Fullerton (United States) [presenting]
Abstract: Although there is a huge literature on feature selection for the Cox model, none of the existing approaches can control the false discovery rate (FDR) unless the sample size tends to infinity. In addition, there is no formal power analysis of the knockoffs framework for survival data in the literature. To address those issues, a novel controlled feature selection approach is proposed using knockoffs for the Cox model. The proposed method is established to enjoy the FDR control in finite samples regardless of the number of covariates. Moreover, under mild regularity conditions, the power of the method is also shown to be asymptotically one as the sample size tends to infinity. To the best of knowledge, this is the first formal theoretical result on the power of the knockoffs procedure in the survival setting. Simulation studies confirm that the method has appealing finite-sample performance with desired FDR control and high power. The performance of the method is further demonstrated through a real data example.
