LOCATION:   1252 Howe Hall
                                     
SPEAKER:   Wes Johnson, University of California-Davis

 

TITLE: Bayesian Semi-parametric Models for Survival Data with Time
Dependent Covariates

 

ABSTRACT:

We discuss three models for survival data with time dependent covariates (TDC), one by Cox and Oakes (1984), the second by Cox (1972), and a new model for TDC's that parallels Cox's original construction, only for the Accelerated Failure Time Model (AFT).  We develop a Markov chain Monte Carlo approach based on modeling the baseline survival distribution as a mixture of Polya Tree processes for each model.  An illustration is presented using Cerebral Edema data that was collected by Drs. Glaser and Kupperman from the UCD School of Medicine. COFFEE: 3:45 p.m., 104 Snedecor Hall