|
Activity Number:
|
360
|
|
Type:
|
Topic Contributed
|
|
Date/Time:
|
Tuesday, August 4, 2009 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #305872 |
|
Title:
|
Some Extensions of the Polya Urn Scheme with Bayesian Applications
|
|
Author(s):
|
Lorenzo Trippa*+
|
|
Companies:
|
The University of Texas M.D. Anderson Cancer Center
|
|
Address:
|
, , ,
|
|
Keywords:
|
bayesian nonparametric models ; Urn schemes ; surival analysis ; partial exchangeability
|
|
Abstract:
|
The reinforcement concept plays a key role in Bayesian statistics. In particular the Polya urn scheme has been extensively studied in Bayesian nonparametrics. We characterize some extensions of the Polya urn scheme. The main purpose of the extensions is that of modeling partially exchangeable populations. The characterizations are illustrated following two guiding principles: the need of specifying prior distributions with clear interpretations and the purpose of constructing flexible Bayesian models with large supports. A fully nonparametric urn-based model for time to event data is specified. Some comparisons between the proposed model and alternative semiparametric approaches are given.
|
