Bayesian Experimental Design and Shannon Information (1997)
| Venue: | In 1997 Proceedings of the Section on Bayesian Statistical Science |
| Citations: | 8 - 0 self |
BibTeX
@INPROCEEDINGS{Sebastiani97bayesianexperimental,
author = {Paola Sebastiani and Henry P. Wynn},
title = {Bayesian Experimental Design and Shannon Information},
booktitle = {In 1997 Proceedings of the Section on Bayesian Statistical Science},
year = {1997},
pages = {176--181}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
The information theoretic approach to optimal design of experiments yields a simple design criterion: the optimal design minimizes the expected posterior entropy of the parameters. Unfortunately, this strategy is often computational infeasible for non-linear problems and numerical approximations are required. This paper reviews the information theoretic approach to design of experiments, and examines computational issues related to the minimization of the expected posterior entropy and its asymptotic approximations. It is shown that Maximum Entropy Sampling simplifies both the formulation of the design criterion and the optimization problem. Numerical advantages are shown in an example, where the exact solution is compared to asymptotic optimal solution.
Keyphrases
bayesian experimental design shannon information information theoretic approach optimal design exact solution design criterion computational issue posterior entropy simple design criterion maximum entropy sampling asymptotic optimal solution numerical advantage non-linear problem numerical approximation expected posterior entropy experiment yield optimization problem asymptotic approximation
