Activity Number:
|
201
|
Type:
|
Topic Contributed
|
Date/Time:
|
Tuesday, August 13, 2002 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Statistical Graphics*
|
Abstract - #301992 |
Title:
|
Benchmark Estimation for Markov Chain Monte Carlo Samples
|
Author(s):
|
Subharup Guha*+ and Steven MacEachern and Mario Peruggia
|
Affiliation(s):
|
Ohio State University and Ohio State University and Ohio State University
|
Address:
|
1958 Neil Ave, Columbus, OH, 43210-1247, USA
|
Keywords:
|
variance reduction ; subsample ; maximum entropy ; stratification
|
Abstract:
|
While studying various features of the posterior distribution of a vector-valued parameter using an MCMC sample, a subsample is often all that is available for analysis. The goal of benchmark estimation is to use the best available information--i.e., the full MCMC sample, to improve future estimates made on the basis of the subsample. We discuss a simple approach to do this. The methodology and benefits of benchmark estimation are illustrated using a well-known example from the literature. We obtain as much as an 80% reduction in MSE with the technique based on a 1-in-10 subsample. Greater benefits accrue with the thinner subsamples often used in practice. Further motivation, results, and investigations will be discussed.
|