A frequentist approach to computer model calibration

Raymond K W Wong, Curtis Storlie, Thomas C M Lee

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

The paper considers the computer model calibration problem and provides a general frequentist solution. Under the framework proposed, the data model is semiparametric with a non-parametric discrepancy function which accounts for any discrepancy between physical reality and the computer model. In an attempt to solve a fundamentally important (but often ignored) identifiability issue between the computer model parameters and the discrepancy function, the paper proposes a new and identifiable parameterization of the calibration problem. It also develops a two-step procedure for estimating all the relevant quantities under the new parameterization. This estimation procedure is shown to enjoy excellent rates of convergence and can be straightforwardly implemented with existing software. For uncertainty quantification, bootstrapping is adopted to construct confidence regions for the quantities of interest. The practical performance of the methodology is illustrated through simulation examples and an application to a computational fluid dynamics model.

Original languageEnglish (US)
JournalJournal of the Royal Statistical Society. Series B: Statistical Methodology
DOIs
StateAccepted/In press - 2016
Externally publishedYes

Fingerprint

Model Calibration
Computer Model
Discrepancy
Parameterization
Uncertainty Quantification
Confidence Region
Bootstrapping
Identifiability
Fluid Model
Computational Fluid Dynamics
General Solution
Data Model
Dynamic Model
Rate of Convergence
Calibration
Software
Methodology
Model calibration
Simulation

Keywords

  • Bootstrap
  • Inverse problem
  • Model misspecification
  • Semiparametric modelling
  • Surrogate model
  • Uncertainty analysis

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

A frequentist approach to computer model calibration. / Wong, Raymond K W; Storlie, Curtis; Lee, Thomas C M.

In: Journal of the Royal Statistical Society. Series B: Statistical Methodology, 2016.

Research output: Contribution to journalArticle

@article{eb0373ffbeb7451cabf64d338b930ae8,
title = "A frequentist approach to computer model calibration",
abstract = "The paper considers the computer model calibration problem and provides a general frequentist solution. Under the framework proposed, the data model is semiparametric with a non-parametric discrepancy function which accounts for any discrepancy between physical reality and the computer model. In an attempt to solve a fundamentally important (but often ignored) identifiability issue between the computer model parameters and the discrepancy function, the paper proposes a new and identifiable parameterization of the calibration problem. It also develops a two-step procedure for estimating all the relevant quantities under the new parameterization. This estimation procedure is shown to enjoy excellent rates of convergence and can be straightforwardly implemented with existing software. For uncertainty quantification, bootstrapping is adopted to construct confidence regions for the quantities of interest. The practical performance of the methodology is illustrated through simulation examples and an application to a computational fluid dynamics model.",
keywords = "Bootstrap, Inverse problem, Model misspecification, Semiparametric modelling, Surrogate model, Uncertainty analysis",
author = "Wong, {Raymond K W} and Curtis Storlie and Lee, {Thomas C M}",
year = "2016",
doi = "10.1111/rssb.12182",
language = "English (US)",
journal = "Journal of the Royal Statistical Society. Series B: Statistical Methodology",
issn = "1369-7412",
publisher = "Wiley-Blackwell",

}

TY - JOUR

T1 - A frequentist approach to computer model calibration

AU - Wong, Raymond K W

AU - Storlie, Curtis

AU - Lee, Thomas C M

PY - 2016

Y1 - 2016

N2 - The paper considers the computer model calibration problem and provides a general frequentist solution. Under the framework proposed, the data model is semiparametric with a non-parametric discrepancy function which accounts for any discrepancy between physical reality and the computer model. In an attempt to solve a fundamentally important (but often ignored) identifiability issue between the computer model parameters and the discrepancy function, the paper proposes a new and identifiable parameterization of the calibration problem. It also develops a two-step procedure for estimating all the relevant quantities under the new parameterization. This estimation procedure is shown to enjoy excellent rates of convergence and can be straightforwardly implemented with existing software. For uncertainty quantification, bootstrapping is adopted to construct confidence regions for the quantities of interest. The practical performance of the methodology is illustrated through simulation examples and an application to a computational fluid dynamics model.

AB - The paper considers the computer model calibration problem and provides a general frequentist solution. Under the framework proposed, the data model is semiparametric with a non-parametric discrepancy function which accounts for any discrepancy between physical reality and the computer model. In an attempt to solve a fundamentally important (but often ignored) identifiability issue between the computer model parameters and the discrepancy function, the paper proposes a new and identifiable parameterization of the calibration problem. It also develops a two-step procedure for estimating all the relevant quantities under the new parameterization. This estimation procedure is shown to enjoy excellent rates of convergence and can be straightforwardly implemented with existing software. For uncertainty quantification, bootstrapping is adopted to construct confidence regions for the quantities of interest. The practical performance of the methodology is illustrated through simulation examples and an application to a computational fluid dynamics model.

KW - Bootstrap

KW - Inverse problem

KW - Model misspecification

KW - Semiparametric modelling

KW - Surrogate model

KW - Uncertainty analysis

UR - http://www.scopus.com/inward/record.url?scp=84965167574&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84965167574&partnerID=8YFLogxK

U2 - 10.1111/rssb.12182

DO - 10.1111/rssb.12182

M3 - Article

AN - SCOPUS:84965167574

JO - Journal of the Royal Statistical Society. Series B: Statistical Methodology

JF - Journal of the Royal Statistical Society. Series B: Statistical Methodology

SN - 1369-7412

ER -