Gaussian process emulation of dynamic computer codes

S. Conti; J. P. Gosling; J. E. Oakley; A. O'Hagan

doi:10.1093/biomet/asp028

Gaussian process emulation of dynamic computer codes

S. Conti^*, J. P. Gosling, J. E. Oakley, A. O'Hagan

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

149 Citations (Scopus)

Abstract

Computer codes are used in scientific research to study and predict the behaviour of complex systems. Their run times often make uncertainty and sensitivity analyses impractical because of the thousands of runs that are conventionally required, so efficient techniques have been developed based on a statistical representation of the code. The approach is less straightforward for dynamic codes, which represent time-evolving systems. We develop a novel iterative system to build a statistical model of dynamic computer codes, which is demonstrated on a rainfall-runoff simulator.

Original language	English
Pages (from-to)	663-676
Number of pages	14
Journal	Biometrika
Volume	96
Issue number	3
DOIs	https://doi.org/10.1093/biomet/asp028
Publication status	Published - Sept 2009

Bibliographical note

Funding Information:
The work in this paper is part of the activities of the Managing Uncertainty in Complex Models project that is funded by a Research Councils UK grant. We thank Peter Reichert for providing the details of the rainfall-runoff model analyzed in this paper. We would also like to thank Professor D. M. Titterington and the anonymous referees for their helpful and stimulating comments on earlier drafts of this paper.

Copyright:
Copyright 2009 Elsevier B.V., All rights reserved.

Keywords

Bayesian inference
Computer experiment
Dynamic simulator
Emulation
Gaussian process
Iterative modelling

Access to Document

10.1093/biomet/asp028

Cite this

@article{4982fc2fdf4b4930ac3906bb606c5f6b,

title = "Gaussian process emulation of dynamic computer codes",

abstract = "Computer codes are used in scientific research to study and predict the behaviour of complex systems. Their run times often make uncertainty and sensitivity analyses impractical because of the thousands of runs that are conventionally required, so efficient techniques have been developed based on a statistical representation of the code. The approach is less straightforward for dynamic codes, which represent time-evolving systems. We develop a novel iterative system to build a statistical model of dynamic computer codes, which is demonstrated on a rainfall-runoff simulator.",

keywords = "Bayesian inference, Computer experiment, Dynamic simulator, Emulation, Gaussian process, Iterative modelling",

author = "S. Conti and Gosling, {J. P.} and Oakley, {J. E.} and A. O'Hagan",

note = "Funding Information: The work in this paper is part of the activities of the Managing Uncertainty in Complex Models project that is funded by a Research Councils UK grant. We thank Peter Reichert for providing the details of the rainfall-runoff model analyzed in this paper. We would also like to thank Professor D. M. Titterington and the anonymous referees for their helpful and stimulating comments on earlier drafts of this paper. Copyright: Copyright 2009 Elsevier B.V., All rights reserved.",

year = "2009",

month = sep,

doi = "10.1093/biomet/asp028",

language = "English",

volume = "96",

pages = "663--676",

journal = "Biometrika",

issn = "0006-3444",

publisher = "Oxford University Press",

number = "3",

}

TY - JOUR

T1 - Gaussian process emulation of dynamic computer codes

AU - Conti, S.

AU - Gosling, J. P.

AU - Oakley, J. E.

AU - O'Hagan, A.

N1 - Funding Information: The work in this paper is part of the activities of the Managing Uncertainty in Complex Models project that is funded by a Research Councils UK grant. We thank Peter Reichert for providing the details of the rainfall-runoff model analyzed in this paper. We would also like to thank Professor D. M. Titterington and the anonymous referees for their helpful and stimulating comments on earlier drafts of this paper. Copyright: Copyright 2009 Elsevier B.V., All rights reserved.

PY - 2009/9

Y1 - 2009/9

N2 - Computer codes are used in scientific research to study and predict the behaviour of complex systems. Their run times often make uncertainty and sensitivity analyses impractical because of the thousands of runs that are conventionally required, so efficient techniques have been developed based on a statistical representation of the code. The approach is less straightforward for dynamic codes, which represent time-evolving systems. We develop a novel iterative system to build a statistical model of dynamic computer codes, which is demonstrated on a rainfall-runoff simulator.

AB - Computer codes are used in scientific research to study and predict the behaviour of complex systems. Their run times often make uncertainty and sensitivity analyses impractical because of the thousands of runs that are conventionally required, so efficient techniques have been developed based on a statistical representation of the code. The approach is less straightforward for dynamic codes, which represent time-evolving systems. We develop a novel iterative system to build a statistical model of dynamic computer codes, which is demonstrated on a rainfall-runoff simulator.

KW - Bayesian inference

KW - Computer experiment

KW - Dynamic simulator

KW - Emulation

KW - Gaussian process

KW - Iterative modelling

UR - http://www.scopus.com/inward/record.url?scp=69249168661&partnerID=8YFLogxK

U2 - 10.1093/biomet/asp028

DO - 10.1093/biomet/asp028

M3 - Article

AN - SCOPUS:69249168661

SN - 0006-3444

VL - 96

SP - 663

EP - 676

JO - Biometrika

JF - Biometrika

IS - 3

ER -

Gaussian process emulation of dynamic computer codes

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this