Optimal Bayesian design for model discrimination via classification
Abstract
Performing optimal Bayesian design for discriminating between competing models is computationally intensive as it involves estimating posterior model probabilities for thousands of simulated datasets. This issue is compounded further when the likelihood functions for the rival models are computationally expensive. A new approach using supervised classification methods is developed to perform Bayesian optimal model discrimination design. This approach requires considerably fewer simulations from the candidate models than previous approaches using approximate Bayesian computation. Further, it is easy to assess the performance of the optimal design through the misclassification error rate. The approach is particularly useful in the presence of models with intractable likelihoods but can also provide computational advantages when the likelihoods are manageable.
Cite
@article{arxiv.1809.05301,
title = {Optimal Bayesian design for model discrimination via classification},
author = {Markus Hainy and David J. Price and Olivier Restif and Christopher Drovandi},
journal= {arXiv preprint arXiv:1809.05301},
year = {2022}
}
Comments
Major revision of previous version: use trees with cross-validation and random forests with out-of-bag predictions to estimate expected loss; training set-based loss estimates are not used anymore; a post-processing step utilising Gaussian process regression is added to the optimisation routine; examples were re-run; extensive reorganisation of contents