Multi-objective optimisation using expected quantile improvement for decision making in disease outbreaks
Multi-objective optimisation using expected quantile improvement for decision making in disease outbreaks
Optimization under uncertainty is important in many applications, particularly to inform policy and decision making in areas such as public health. A key source of uncertainty arises from the incorporation of environmental variables as inputs into computational models or simulators. Such variables represent uncontrollable features of the optimization problem, and reliable decision making must account for the uncertainty they propagate to the simulator outputs. Often, multiple, competing objectives are defined from these outputs such that the final optimal decision is a compromise between different goals. Here, we present emulation-based optimization methodology for such problems that extends expected quantile improvement (EQI) to address multiobjective optimization. Focusing on the practically important case of two objectives, we use a sequential design strategy to identify the Pareto front of optimal solutions. Uncertainty from the environmental variables is integrated out using Monte Carlo samples from the simulator. Interrogation of the expected output from the simulator is facilitated by use of (Gaussian process) emulators. The methodology is demonstrated on an optimization problem from public health involving the dispersion of anthrax spores across a spatial terrain. Environmental variables include meteorological features that impact the dispersion, and the methodology identifies the Pareto front even when there is considerable input uncertainty.
(quantile) expected improvement, emulation, Gaussian processes, stochastic optimization
228-250
Semochkina, Daria
011d4fa0-cf50-4739-890e-7f453027432f
Forrester, Alexander I.J.
a96fd147-f59b-4312-9fca-c59e8f14ed79
Woods, David C.
ae21f7e2-29d9-4f55-98a2-639c5e44c79c
7 March 2025
Semochkina, Daria
011d4fa0-cf50-4739-890e-7f453027432f
Forrester, Alexander I.J.
a96fd147-f59b-4312-9fca-c59e8f14ed79
Woods, David C.
ae21f7e2-29d9-4f55-98a2-639c5e44c79c
Semochkina, Daria, Forrester, Alexander I.J. and Woods, David C.
(2025)
Multi-objective optimisation using expected quantile improvement for decision making in disease outbreaks.
SIAM/ASA Journal on Uncertainty Quantification, 13 (1), .
(doi:10.1137/24M1633625).
Abstract
Optimization under uncertainty is important in many applications, particularly to inform policy and decision making in areas such as public health. A key source of uncertainty arises from the incorporation of environmental variables as inputs into computational models or simulators. Such variables represent uncontrollable features of the optimization problem, and reliable decision making must account for the uncertainty they propagate to the simulator outputs. Often, multiple, competing objectives are defined from these outputs such that the final optimal decision is a compromise between different goals. Here, we present emulation-based optimization methodology for such problems that extends expected quantile improvement (EQI) to address multiobjective optimization. Focusing on the practically important case of two objectives, we use a sequential design strategy to identify the Pareto front of optimal solutions. Uncertainty from the environmental variables is integrated out using Monte Carlo samples from the simulator. Interrogation of the expected output from the simulator is facilitated by use of (Gaussian process) emulators. The methodology is demonstrated on an optimization problem from public health involving the dispersion of anthrax spores across a spatial terrain. Environmental variables include meteorological features that impact the dispersion, and the methodology identifies the Pareto front even when there is considerable input uncertainty.
Text
Multi_objective_optimisation_using_expected_quantile_improvement_for_decision_making_in_disease_outbreaks_-6
- Accepted Manuscript
More information
Accepted/In Press date: 17 October 2024
Published date: 7 March 2025
Keywords:
(quantile) expected improvement, emulation, Gaussian processes, stochastic optimization
Identifiers
Local EPrints ID: 496493
URI: http://eprints.soton.ac.uk/id/eprint/496493
PURE UUID: 8d8f41a4-5d58-43e3-a7bb-316f2c2df90f
Catalogue record
Date deposited: 17 Dec 2024 17:33
Last modified: 28 Aug 2025 04:01
Export record
Altmetrics
Contributors
Author:
Alexander I.J. Forrester
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics