|
Activity Number:
|
486
|
|
Type:
|
Invited
|
|
Date/Time:
|
Thursday, August 2, 2007 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #308028 |
|
Title:
|
Bayesian Semiparametric Modeling Based on Mixtures of Polya Trees
|
|
Author(s):
|
Timothy E. Hanson*+
|
|
Companies:
|
The University of Minnesota
|
|
Address:
|
Div. of Biostatistics, A460 Mayo Building, MMC 303, Minneapolis, MN, 55455,
|
|
Keywords:
|
Nonparametric ; mixed model
|
|
Abstract:
|
Relative to Dirichlet process mixtures (DPM) and other nonparametric prior processes, the Polya tree and mixtures of Polya tree (MPT) priors have seen relatively limited use. In this talk I will briefly provide some background on MPTs and implementation notes, then move on to a number of real data examples where the MPT model improves prediction relative to a parametric model or equivalent DPM model. Examples will include survival models with time dependent longitudinal trajectories, survival models with spatial frailties, diagnostic serology test data modeled subject to a stochastic order constraint, dependent Polya tree processes, and Polya tree mixtures for meta-analysis in non-normal settings.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2007 program |
