This is the program for the 2010 Joint Statistical Meetings in Vancouver, British Columbia.

Abstract Details

Activity Number: 490
Type: Invited
Date/Time: Wednesday, August 4, 2010 : 10:30 AM to 12:20 PM
Sponsor: IMS
Abstract - #306274
Title: Discovering Regression Structure with a Bayesian Ensemble
Author(s): Edward I. George*+
Companies: University of Pennsylvania
Address: 3730 Walnut St, Philadelphia, PA, 19104,
Keywords: boosting ; model building ; nonparametric regression ; random forests ; trees ; variable selection

A Bayesian ensemble approach can be used to discover and learn about an unknown regression relationship between a variable of interest y and a vector of predictors x. The basic idea is to model the conditional distribution of y given x by a sum of random basis elements plus a flexible noise distribution. BART, a special case which uses random regression trees as basis functions, can automatically produce the predictive distribution of y at any x (in or out of sample). It can do this for nonlinear relationships, even when hidden within a large number of irrelevant predictors, and by constraining the number of trees to create a bottleneck effect, it can be used for model free variable selection. Ultimately, the many features of such an approach may be seen as a valuable first step towards model building for high dimensional data. This is joint work with H. Chipman and R. McCulloch.

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 2010 program

2010 JSM Online Program Home

For information, contact or phone (888) 231-3473.

If you have questions about the Continuing Education program, please contact the Education Department.