Small Sample Bayesian Designs for Complex High-Dimensional Models Based on Information Gained Using Fast Approximations

Cumming, J.A.; Goldstein, M.

Abstract

We consider the problem of designing for complex high-dimensional computer models that can be evaluated at different levels of accuracy. Ordinarily, this requires performing many expensive evaluations of the most accurate version of the computer model to obtain a reasonable coverage of the design space. In some cases, it is possible to supplement the information from the accurate model evaluations with a large number of evaluations of a cheap, approximate version of the computer model to enable a more informed design choice. We describe an approach that combines the information from both the approximate model and the accurate model into a single multiscale emulator for the computer model. We then propose a design strategy for selecting a small number of expensive evaluations of the accurate computer model based on our multiscale emulator and a decomposition of the input parameter space. We illustrate our methodology with an example concerning a computer simulation of a hydrocarbon reservoir.

Citation

Cumming, J., & Goldstein, M. (2009). Small Sample Bayesian Designs for Complex High-Dimensional Models Based on Information Gained Using Fast Approximations. Technometrics, 51(4), 377-388. https://doi.org/10.1198/tech.2009.08015

• Journal Article Type: Article
• Publication Date: Nov 1, 2009
• Deposit Date: Feb 25, 2011
• Publicly Available Date: Aug 8, 2016
• Journal: Technometrics
• Print ISSN: 0040-1706
• Electronic ISSN: 1537-2723
• Publisher: American Statistical Association
• Peer Reviewed: Peer Reviewed
• Volume: 51
• Issue: 4
• Pages: 377-388
• DOI: https://doi.org/10.1198/tech.2009.08015
• Public URL: https://durham-repository.worktribe.com/output/1543206