2003/09/01 by Kenneth J. Ryan · 176 citations
Decision Sciences · Computer Science · Mathematics · #Optimal Experimental Design Methods #Probabilistic and Robust Engineering Design #Advanced Multi-Objective Optimization Algorithms #Markov chain Monte Carlo #Estimator #Bayesian probability #Monte Carlo method #Limit (mathematics) #Computer science #Mathematical optimization #Markov chain #Mathematics #Machine learning #Statistics #Artificial intelligence
paper · doi:10.1198/1061860032012
published in Journal of Computational and Graphical Statistics 12(3), 585-603 (Taylor & Francis)
openalex publication_date 2003/09/01 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/01
Expected gain in Shannon information is commonly suggested as a Bayesian design evaluation criterion. Because estimating expected information gains is computationally expensive, examples in which they have been successfully used in identifying Bayes optimal designs are both few and typically quite simplistic. This article discusses in general some properties of estimators of expected information gains based on Markov chain Monte Carlo (MCMC) and Laplacian approximations. We then investigate some issues that arise when applying these methods to the problem of experimental design in the (technically nontrivial) random fatigue-limit model of Pascual and Meeker. An example comparing follow-up designs for a laminate panel study is provided.