Frequentist history matching with interval predictor models

Jonathan Sadeghi, Marco De Angelis, Edoardo Patelli

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

In this paper a novel approach is presented for history matching models without making assumptions about the measurement error. Interval Predictor Models are used to robustly model the observed data and hence a novel figure of merit is proposed to quantify the quality of matches in a frequentist probabilistic framework. The proposed method yields bounds on the p-values from frequentist inference. The method is first applied to a simple example and then to a realistic case study (the Imperial College Fault Model) in order to evaluate its applicability and efficacy. When there is no modelling error the method identifies a feasible region for the matched parameters, which for our test case contained the truth case. When attempting to match one model to data from a different model, a region close to the truth case was identified. The effect of increasing the number of data points on the history matching is also discussed.
LanguageEnglish
Pages29-48
Number of pages20
JournalApplied Mathematical Modelling
Volume61
Early online date17 Apr 2018
DOIs
Publication statusPublished - 1 Sep 2018

Fingerprint

History Matching
Predictors
Interval
Model
Modeling Error
Feasible region
p-Value
Measurement errors
Measurement Error
Efficacy
Figure
Quantify
Fault
Evaluate

Keywords

  • interval predictor models
  • history matching
  • surrogate model
  • inverse problem
  • imprecise probability
  • frequentist inference

Cite this

Sadeghi, Jonathan ; Angelis, Marco De ; Patelli, Edoardo. / Frequentist history matching with interval predictor models. In: Applied Mathematical Modelling. 2018 ; Vol. 61. pp. 29-48.
@article{889c0a109f80473d84db8903f02132ab,
title = "Frequentist history matching with interval predictor models",
abstract = "In this paper a novel approach is presented for history matching models without making assumptions about the measurement error. Interval Predictor Models are used to robustly model the observed data and hence a novel figure of merit is proposed to quantify the quality of matches in a frequentist probabilistic framework. The proposed method yields bounds on the p-values from frequentist inference. The method is first applied to a simple example and then to a realistic case study (the Imperial College Fault Model) in order to evaluate its applicability and efficacy. When there is no modelling error the method identifies a feasible region for the matched parameters, which for our test case contained the truth case. When attempting to match one model to data from a different model, a region close to the truth case was identified. The effect of increasing the number of data points on the history matching is also discussed.",
keywords = "interval predictor models, history matching, surrogate model, inverse problem, imprecise probability, frequentist inference",
author = "Jonathan Sadeghi and Angelis, {Marco De} and Edoardo Patelli",
year = "2018",
month = "9",
day = "1",
doi = "10.1016/j.apm.2018.04.003",
language = "English",
volume = "61",
pages = "29--48",
journal = "Applied Mathematical Modelling",
issn = "0307-904X",

}

Frequentist history matching with interval predictor models. / Sadeghi, Jonathan; Angelis, Marco De; Patelli, Edoardo.

In: Applied Mathematical Modelling, Vol. 61, 01.09.2018, p. 29-48.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Frequentist history matching with interval predictor models

AU - Sadeghi, Jonathan

AU - Angelis, Marco De

AU - Patelli, Edoardo

PY - 2018/9/1

Y1 - 2018/9/1

N2 - In this paper a novel approach is presented for history matching models without making assumptions about the measurement error. Interval Predictor Models are used to robustly model the observed data and hence a novel figure of merit is proposed to quantify the quality of matches in a frequentist probabilistic framework. The proposed method yields bounds on the p-values from frequentist inference. The method is first applied to a simple example and then to a realistic case study (the Imperial College Fault Model) in order to evaluate its applicability and efficacy. When there is no modelling error the method identifies a feasible region for the matched parameters, which for our test case contained the truth case. When attempting to match one model to data from a different model, a region close to the truth case was identified. The effect of increasing the number of data points on the history matching is also discussed.

AB - In this paper a novel approach is presented for history matching models without making assumptions about the measurement error. Interval Predictor Models are used to robustly model the observed data and hence a novel figure of merit is proposed to quantify the quality of matches in a frequentist probabilistic framework. The proposed method yields bounds on the p-values from frequentist inference. The method is first applied to a simple example and then to a realistic case study (the Imperial College Fault Model) in order to evaluate its applicability and efficacy. When there is no modelling error the method identifies a feasible region for the matched parameters, which for our test case contained the truth case. When attempting to match one model to data from a different model, a region close to the truth case was identified. The effect of increasing the number of data points on the history matching is also discussed.

KW - interval predictor models

KW - history matching

KW - surrogate model

KW - inverse problem

KW - imprecise probability

KW - frequentist inference

U2 - 10.1016/j.apm.2018.04.003

DO - 10.1016/j.apm.2018.04.003

M3 - Article

VL - 61

SP - 29

EP - 48

JO - Applied Mathematical Modelling

T2 - Applied Mathematical Modelling

JF - Applied Mathematical Modelling

SN - 0307-904X

ER -