Toggle Main Menu Toggle Search

Open Access padlockePrints

Bayesian Experimental Design Without Posterior Calculations: An Adversarial Approach

Lookup NU author(s): Dr Dennis Prangle, Sophie Harbisher, Dr Colin GillespieORCiD

Downloads


Licence

This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).


Abstract

© 2023 International Society for Bayesian Analysis. Most computational approaches to Bayesian experimental design require making posterior calculations repeatedly for a large number of potential designs and/or simulated datasets. This can be expensive and prohibit scaling up these methods to models with many parameters, or designs with many unknowns to select. We introduce an efficient alternative approach without posterior calculations, based on optimising the expected trace of the Fisher information, as discussed by Walker (2016). We illustrate drawbacks of this approach, including lack of invariance to reparameterisation and encouraging designs in which one parameter combination is inferred accurately but not any others. We show these can be avoided by using an adversarial approach: the experimenter must select their design while a critic attempts to select the least favourable parameterisation. We present theoretical properties of this approach and show it can be used with gradient based optimisation methods to find designs efficiently in practice.


Publication metadata

Author(s): Prangle D, Harbisher S, Gillespie CS

Publication type: Article

Publication status: Published

Journal: Bayesian Analysis

Year: 2023

Volume: 18

Issue: 1

Pages: 133-163

Print publication date: 01/03/2023

Online publication date: 08/02/2022

Acceptance date: 02/04/2018

Date deposited: 10/01/2023

ISSN (print): 1936-0975

ISSN (electronic): 1931-6690

Publisher: International Society for Bayesian Analysis

URL: https://doi.org/10.1214/22-BA1306

DOI: 10.1214/22-BA1306


Altmetrics

Altmetrics provided by Altmetric


Funding

Funder referenceFunder name
EPSRC

Share