Abstract #300222

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 #300222
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 - #300222
Title: Efficient Sampling from Nonstandard Distributions Using Neural Network Approximations
Author(s): Lennart Hoogerheide*+ and Herman Koene van Dijk
Companies: Erasmus University Rotterdam and Erasmus University Rotterdam
Address: Tinbergen Institute, Rotterdam, NL-3062 PA, Netherlands
Keywords: Neural networks ; Markov chain Monte Carlo methods ; Importance sampling
Abstract:

The performance of integration methods like importance sampling or Markov chain Monte Carlo procedures greatly depends on the choice of the importance or candidate density. Usually, an importance or candidate density that is 'close' to the target density will efficiently yield reliable sampling results. Neural networks seem to be natural importance and candidate densities, as they have a universal approximation property and are easy to sample from. This sampling can be done either directly or using a Gibbs sampling technique (possibly using auxiliary variables), depending on the exact specification of the neural network. Part of the computing time is invested in constructing a neural network approximation to the target density in order to make the sampling more efficient and/or reliable. Experiments using mixtures of distributions and empirical examples involving some nonstandard, nonelliptical posterior distributions indicate the feasibility of our approach.

  • 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