Bayesian Design of Experiments for Intractable Likelihood Models Using Coupled Auxiliary Models and Multivariate Emulation
Authors:
- Antony OverstallAOAntony Overstall
- James McGreeJMJames McGree
Open Access
- 1 March 2020
- journal article
- research article
- Published by Institute of Mathematical Statistics in Bayesian Analysis
- Vol. 15 (1) , 103-131
Abstract
A Bayesian design is given by maximising an expected utility over a design space. The utility is chosen to represent the aim of the experiment and its expectation is taken with respect to all unknowns: responses, parameters and/or models. Although straightforward in principle, there are several challenges to finding Bayesian designs in practice. Firstly, the utility and expected utility are rarely available in closed form and require approximation. Secondly, the design space can be of high-dimensionality. In the case of intractable likelihood models, these problems are compounded by the fact that the likelihood function, whose evaluation is required to approximate the expected utility, is not available in closed form. A strategy is proposed to find Bayesian designs for intractable likelihood models. It relies on the development of an automatic, auxiliary modelling approach, using multivariate Gaussian process emulators, to approximate the likelihood function. This is then combined with a copula-based approach to approximate the marginal likelihood (a quantity commonly required to evaluate many utility functions). These approximations are demonstrated on examples of stochastic process models involving experimental aims of both parameter estimation and model comparison.Keywords
This publication has 48 references indexed in Scilit:
- Optimal design of measurements on queueing systemsQueueing Systems, 2014
- Towards Bayesian experimental design for nonlinear models that require a large number of sampling timesComputational Statistics & Data Analysis, 2014
- Bayesian Experimental Design for Models with Intractable LikelihoodsBiometrics, 2013
- Optimal design of experimental epidemicsJournal of Statistical Planning and Inference, 2013
- Simulation-based optimal Bayesian experimental design for nonlinear systemsJournal of Computational Physics, 2013
- Pair Copula Constructions for Multivariate Discrete DataJournal of the American Statistical Association, 2012
- Bayesian emulation of complex multi-output and dynamic computer modelsJournal of Statistical Planning and Inference, 2010
- Stochastic modeling of aphid population growth with nonlinear, power-law dynamicsMathematical Biosciences, 2007
- Bayesian-Optimal Design via Interacting Particle SystemsJournal of the American Statistical Association, 2006
- Optimal Bayesian Design by Inhomogeneous Markov Chain SimulationJournal of the American Statistical Association, 2004