207
reads

Hung Hung

2008-07-15
13:30:00 - 15:00:00

Nonparametric Estimation Methods for the Optimal Composite Markers with Censored Survival Data

308 , Mathematics Research Center Building (ori. New Math. Bldg.)

To increase the predictive abilities of several markers on the vital statuses over time, our research interest mainly focuses on seeking appropriate composite markers with the highest time-dependent receiver operating characteristic (ROC) curves. An extended generalized linear model (EGLM) with time-varying coefficients and an unknown bivariate link function is used to characterize the conditional distribution of the time to CAD-related death. Two nonparametric procedures are proposed to estimate the optimal composite markers, linear predictors in the EGLM model. The estimation methods for the accuracy measurements of the optimal composite markers are also proposed. In this study, we establish the theoretical results of the estimators and examine the corresponding finite sample properties through a series of simulations with different sample sizes, censoring rates, and censoring mechanisms. Our optimization procedures and estimators are further shown to be useful through an application to a prospective cohort study of patients undergoing angiography.