Spring 2001 Seminar Series - February 15, 2001
University of Minnesota
School of Statistics
College of Liberal Arts
Proportional Hazards Regression Model with Unknown Link Function and
Applications to Longitudinal Data
Wei Wang
University of California, Davis
(Statistics Search Candidate)
Thursday, February 15, 2001
4:00 PM,
B10
Ford Hall
Minneapolis, East Bank Campus
Social at 3:30 PM,
300
Ford Hall
Abstract
Cox proportional hazards regression model usually assumes that covariates
have log-linear effects on the hazard function. Now we consider a more
general proportional hazards regression model with a nonparametric link
function. We mainly focus on the situation when the baseline hazard
function is also unspecified. A two-step iterative algorithm is proposed
to estimate the link function and the covariate effects. We make inferencebased on a local version of the partial likelihood.
Large sample properties are discussed and several simulation studies are
conducted to evaluate the performance of this estimation procedure.
We apply this method to data with longitudinal covariates and provide a
way to handle possible missing values using functional principal
components analysis.