Current browse context:
stat.CO
Change to browse by:
References & Citations
Statistics > Computation
Title: Unbiased MLMC stochastic gradient-based optimization of Bayesian experimental designs
(Submitted on 18 May 2020 (v1), last revised 28 Jul 2021 (this version, v3))
Abstract: In this paper we propose an efficient stochastic optimization algorithm to search for Bayesian experimental designs such that the expected information gain is maximized. The gradient of the expected information gain with respect to experimental design parameters is given by a nested expectation, for which the standard Monte Carlo method using a fixed number of inner samples yields a biased estimator. In this paper, applying the idea of randomized multilevel Monte Carlo (MLMC) methods, we introduce an unbiased Monte Carlo estimator for the gradient of the expected information gain with finite expected squared $\ell_2$-norm and finite expected computational cost per sample. Our unbiased estimator can be combined well with stochastic gradient descent algorithms, which results in our proposal of an optimization algorithm to search for an optimal Bayesian experimental design. Numerical experiments confirm that our proposed algorithm works well not only for a simple test problem but also for a more realistic pharmacokinetic problem.
Submission history
From: Takashi Goda [view email][v1] Mon, 18 May 2020 01:02:31 GMT (5270kb)
[v2] Sat, 9 Jan 2021 02:21:46 GMT (296kb,D)
[v3] Wed, 28 Jul 2021 05:13:20 GMT (298kb,D)
Link back to: arXiv, form interface, contact.