Abstract:We address the computational efficiency in solving the A-optimal Bayesian design of experiments problems for which the observational map is based on partial differential equations and, consequently, is computationally expensive to evaluate. A-optimality is a widely used and easy-to-interpret criterion for Bayesian experimental design. This criterion seeks the optimal experimental design by minimizing the expected conditional variance, which is also known as the expected posterior variance. This study presents a novel likelihood-free approach to the A-optimal experimental design that does not require sampling or integrating the Bayesian posterior distribution. The expected conditional variance is obtained via the variance of the conditional expectation using the law of total variance, and we take advantage of the orthogonal projection property to approximate the conditional expectation. We derive an asymptotic error estimation for the proposed estimator of the expected conditional variance and show that the intractability of the posterior distribution does not affect the performance of our approach. We use an artificial neural network (ANN) to approximate the nonlinear conditional expectation in the implementation of our method. We then extend our approach for dealing with the case that the domain of experimental design parameters is continuous by integrating the training process of the ANN into minimizing the expected conditional variance. Through numerical experiments, we demonstrate that our method greatly reduces the number of observation model evaluations compared with widely used importance sampling-based approaches. This reduction is crucial, considering the high computational cost of the observational models. Code is available at <a class="link-external link-https" href="https://github.com/vinh-tr-hoang/DOEviaPACE" rel="external noopener nofollow">this https URL</a>.

Computing optimal experimental designs with respect to a compound Bayes risk criterion

Optimal Designs for Minimax-Criteria in Random Coefficient Regression Models

The polytope of optimal approximate designs: extending the selection of informative experiments

An approach for finding fully Bayesian optimal designs using normal-based approximations to loss functions

Computing Optimal Designs of multiresponse Experiments reduces to Second-Order Cone Programming

Efficient computation of Bayesian optimal discriminating designs

Scalable method for Bayesian experimental design without integrating over posterior distribution

Nonlinear Bayesian optimal experimental design using logarithmic Sobolev inequalities

Comparing robustness properties of optimal designs under standard and compound criteria

Heuristic construction of exact experimental designs under multiple resource constraints

Laplace-based strategies for Bayesian optimal experimental design with nuisance uncertainty

Efficient construction of Bayes optimal designs for stochastic process models

An Induced Natural Selection Heuristic for Finding Optimal Bayesian Experimental Designs

Optimal designs for frequentist model averaging

Efficient Computational Algorithm for Optimal Continuous Experimental Designs

An Adaptive Algorithm based on High-Dimensional Function Approximation to obtain Optimal Designs

Optimal Bayesian design for model discrimination via classification

Accelerating Experimental Design by Incorporating Experimenter Hunches

Compound Optimum Designs for Clinical Trials in Personalized Medicine

Equivalence theorems for compound design problems with application in mixed models

Likelihood-free Simulation-based Optimal Design