|
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 - #308790 |
|
Title:
|
Nonparametric Bayes Local Regression and Variable Selection
|
|
Author(s):
|
Yeonseung Chung*+ and David B. Dunson
|
|
Companies:
|
The University of North Carolina at Chapel Hill and National Institute of Environmental Health Sciences
|
|
Address:
|
222 Standish Dr, Chapel Hill, NC, 27517,
|
|
Keywords:
|
|
|
Abstract:
|
Flexibly characterizing the relationship between a response and multiple predictors has been a great interest in many applications. In such settings, interest focuses on estimating predictor-dependent distributions, while also identifying significant predictors globally and within a local region. To address these two problems simultaneously, we propose a new class of stick-breaking prior, called local Dirichlet process (lDP), for the collection of dependent distributions. A hierarchical variable selection mixture structure is incorporated in the base measure to allow uncertainty for the predictors to be included. A blocked Gibbs sampler stochastic search algorithm for the local Dirichlet process mixture (lDPM) is proposed for posterior computation. Theoretical properties are discussed and the methods are illustrated using simulated examples and an epidemiologic application.
|
- 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 |
