Emulation of utility functions over a set of permutations: sequencing reliability growth tasks

Kevin J Wilson, Daniel A Henderson, John Quigley

Research output: Contribution to journalArticle

Abstract

We consider Bayesian design of experiments problems in which we maximise the prior expectation of a utility function over a set of permutations, for example when sequencing a number of tasks to perform. When the number of tasks is large and the expected utility is expensive to compute, it may be unreasonable or infeasible to evaluate the expected utility of all permutations. We propose an approach to emulate the expected utility using a surrogate function based on a parametric probabilistic model for permutations. The surrogate function is fitted by maximising the correlation with the expected utility over a set of training points. We propose a suitable transformation of the expected utility to improve the fit. We provide results linking the correlation between the two functions and the number of expected utility evaluations to undertake. The approach is applied to the sequencing of reliability growth tasks in the development of hardware systems, in which there is a large number of potential tasks to perform and engineers are interested in meeting a reliability target subject to minimising costs and time. An illustrative example shows how the approach can be used and a simulation study demonstrates the performance of the approach more generally.
LanguageEnglish
Number of pages28
JournalTechnometrics
Early online date12 Sep 2017
DOIs
Publication statusE-pub ahead of print - 12 Sep 2017

Fingerprint

Reliability Growth
Emulation
Expected Utility
Utility Function
Sequencing
Permutation
Bayesian Design
Design of experiments
Design of Experiments
Parametric Model
Hardware
Engineers
Probabilistic Model
Linking
Maximise
Simulation Study
Costs
Target
Evaluate
Evaluation

Keywords

  • multi-attribute utility
  • Bayesian design of experiments
  • design for reliability
  • Benter model

Cite this

@article{2ba49e5c0925446298197760729074e9,
title = "Emulation of utility functions over a set of permutations: sequencing reliability growth tasks",
abstract = "We consider Bayesian design of experiments problems in which we maximise the prior expectation of a utility function over a set of permutations, for example when sequencing a number of tasks to perform. When the number of tasks is large and the expected utility is expensive to compute, it may be unreasonable or infeasible to evaluate the expected utility of all permutations. We propose an approach to emulate the expected utility using a surrogate function based on a parametric probabilistic model for permutations. The surrogate function is fitted by maximising the correlation with the expected utility over a set of training points. We propose a suitable transformation of the expected utility to improve the fit. We provide results linking the correlation between the two functions and the number of expected utility evaluations to undertake. The approach is applied to the sequencing of reliability growth tasks in the development of hardware systems, in which there is a large number of potential tasks to perform and engineers are interested in meeting a reliability target subject to minimising costs and time. An illustrative example shows how the approach can be used and a simulation study demonstrates the performance of the approach more generally.",
keywords = "multi-attribute utility, Bayesian design of experiments, design for reliability, Benter model",
author = "Wilson, {Kevin J} and Henderson, {Daniel A} and John Quigley",
year = "2017",
month = "9",
day = "12",
doi = "10.1080/00401706.2017.1377637",
language = "English",
journal = "Technometrics",
issn = "0040-1706",

}

Emulation of utility functions over a set of permutations : sequencing reliability growth tasks. / Wilson, Kevin J; Henderson, Daniel A; Quigley, John.

In: Technometrics, 12.09.2017.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Emulation of utility functions over a set of permutations

T2 - Technometrics

AU - Wilson, Kevin J

AU - Henderson, Daniel A

AU - Quigley, John

PY - 2017/9/12

Y1 - 2017/9/12

N2 - We consider Bayesian design of experiments problems in which we maximise the prior expectation of a utility function over a set of permutations, for example when sequencing a number of tasks to perform. When the number of tasks is large and the expected utility is expensive to compute, it may be unreasonable or infeasible to evaluate the expected utility of all permutations. We propose an approach to emulate the expected utility using a surrogate function based on a parametric probabilistic model for permutations. The surrogate function is fitted by maximising the correlation with the expected utility over a set of training points. We propose a suitable transformation of the expected utility to improve the fit. We provide results linking the correlation between the two functions and the number of expected utility evaluations to undertake. The approach is applied to the sequencing of reliability growth tasks in the development of hardware systems, in which there is a large number of potential tasks to perform and engineers are interested in meeting a reliability target subject to minimising costs and time. An illustrative example shows how the approach can be used and a simulation study demonstrates the performance of the approach more generally.

AB - We consider Bayesian design of experiments problems in which we maximise the prior expectation of a utility function over a set of permutations, for example when sequencing a number of tasks to perform. When the number of tasks is large and the expected utility is expensive to compute, it may be unreasonable or infeasible to evaluate the expected utility of all permutations. We propose an approach to emulate the expected utility using a surrogate function based on a parametric probabilistic model for permutations. The surrogate function is fitted by maximising the correlation with the expected utility over a set of training points. We propose a suitable transformation of the expected utility to improve the fit. We provide results linking the correlation between the two functions and the number of expected utility evaluations to undertake. The approach is applied to the sequencing of reliability growth tasks in the development of hardware systems, in which there is a large number of potential tasks to perform and engineers are interested in meeting a reliability target subject to minimising costs and time. An illustrative example shows how the approach can be used and a simulation study demonstrates the performance of the approach more generally.

KW - multi-attribute utility

KW - Bayesian design of experiments

KW - design for reliability

KW - Benter model

U2 - 10.1080/00401706.2017.1377637

DO - 10.1080/00401706.2017.1377637

M3 - Article

JO - Technometrics

JF - Technometrics

SN - 0040-1706

ER -