|
Activity Number:
|
96
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Monday, July 30, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #308343 |
|
Title:
|
Bayesian Variable Selection in Gaussian Process for Cox Models
|
|
Author(s):
|
Naijun Sha*+ and Marina Vannucci and Mahlet G. Tadesse
|
|
Companies:
|
University of Texas at El Paso and Texas A&M University and University of Pennsylvania
|
|
Address:
|
500 W University Ave, El Paso, TX, 79968,
|
|
Keywords:
|
Bayesian variable selection ; Gaussian Process ; Cox Model ; Latent Variable ; Censored Time ; MCMC
|
|
Abstract:
|
In this paper, we investigate variable selection methods for Cox's proportional hazard model via a Gaussian process. We develop selection methods that allow for censored data. Our methods lead to simultaneously estimates of the survival function as well as to the identification of the factors that affect the survival outcome. We handle the problem of selecting a few predictors among the prohibitively vast number of variables through the introduction of a binary exclusion/inclusion latent vector. This vector is updated via an MCMC technique to identify promising models. We describe strategies for posterior inference and explore the performance of the methodology with simulated and real datasets.
|
