Abstract #301718

This is the preliminary program for the 2003 Joint Statistical Meetings in San Francisco, California. Currently included in this program is the "technical" program, schedule of invited, topic contributed, regular contributed and poster sessions; Continuing Education courses (August 2-5, 2003); and Committee and Business Meetings. This on-line program will be updated frequently to reflect the most current revisions.

To View the Program:
You may choose to view all activities of the program or just parts of it at any one time. All activities are arranged by date and time.

The views expressed here are those of the individual authors
and not necessarily those of the ASA or its board, officers, or staff.

Back to main JSM 2003 Program page
JSM 2003 Abstract #301718
Activity Number: 405
Type: Topic Contributed
Date/Time: Wednesday, August 6, 2003 : 2:00 PM to 3:50 PM
Sponsor: Section on Bayesian Stat. Sciences
Abstract - #301718
Title: Parsimonious Markov Chain Monte Carlo for Models with Latent Variables
Author(s): Silvia Fruehwirth-Schnatter*+
Companies: Johannes Kepler Universität
Address: Altenbergerstasse 69, Linz, , A-4040, Austria
Keywords: hierarchical models ; covariance selection ; Markov chain Monte Carlo ; variable selection
Abstract:

A well-known problem for complex models involving latent variables is slow convergence of straightforward MCMC schemes. A common cause is poor parameterization of the latent structure in combination with models that are overparameterized in light of the data. A typical example would be a random-effects model with some of the random effects being nearly deterministic. We consider parsimonious MCMC methods where sampling the unknown model parameters is carried out jointly with finding a parsimonious representation of the underlying model structure. To this aim, suitable selection variables are introduced that are sampled jointly with the parameters. Details will be presented for hierarchal random-effects models. Model flexibility is introduced both for the mean structure, as in common Bayesian variable selection, as well as in the covariance structure of the latent process. Joint covariance selection in combination with choosing the right parameterization of the latent process is shown to be an efficient computational tool for carrying out MCMC.

  • 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 2003 program

JSM 2003 For information, contact meetings@amstat.org or phone (703) 684-1221. If you have questions about the Continuing Education program, please contact the Education Department.
Revised March 2003