Seminars: Dept Seminar

Censored quantile regression

Abstract:

Quantile regression provides a valuable supplement to Cox proportional hazards model for analyzing survival data where censoring is common. Existing methods for censored quantile regression often require stringent assumptions such as linearity of all quantile functions, which restrict model flexibility and complicate computation. In this talk, I will present an index-based estimation method for censored quantile regression to accommodate high dimensional covariates. The new approach adopts the redistribution-of-mass idea and employs a local reweighting scheme. Its validity only requires conditional independence of the survival time and the censoring variable given the covariates, and linearity at the particular quantile level of interest. In addition, I will discuss penalization methods for variable selection in censored quantile regression.