Merging expert and empirical data for rare event frequency estimation: pool homogenisation for empirical Bayes models

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Empirical Bayes provides one approach to estimating the frequency of rare events as a weighted average of the frequencies of an event and a pool of events. The pool will draw upon, for example, events with similar precursors. The higher the degree of homogeneity of the pool, then the Empirical Bayes estimator will be more accurate. We propose and evaluate a new method using homogenisation factors under the assumption that events are generated from a Homogeneous Poisson Process. The homogenisation factors are scaling constants, which can be elicited through structured expert judgement and used to align the frequencies of different events, hence homogenising the pool. The estimation error relative to the homogeneity of the pool is examined theoretically indicating that reduced error is associated with larger pool homogeneity. The effects of misspecified expert assessments of the homogenisation factors are examined theoretically and through simulation experiments. Our results show that the proposed Empirical Bayes method using homogenisation factors is robust under different degrees of misspecification.
LanguageEnglish
Pages687-695
Number of pages9
JournalReliability Engineering and System Safety
Volume96
Issue number6
Early online date5 Jan 2011
DOIs
Publication statusPublished - Jun 2011

Fingerprint

Homogenization method
Frequency Estimation
Frequency estimation
Empirical Bayes
Rare Events
Merging
Homogenization
Homogeneity
Error analysis
Empirical Bayes Method
Empirical Bayes Estimator
Expert Judgment
Homogenization Method
Model
Misspecification
Weighted Average
Estimation Error
Poisson process
Precursor
Experiments

Keywords

  • poisson processes
  • empirical bayes
  • pairwise comparison
  • PRA

Cite this

@article{332aae0a089f4d458b08561249b823a5,
title = "Merging expert and empirical data for rare event frequency estimation: pool homogenisation for empirical Bayes models",
abstract = "Empirical Bayes provides one approach to estimating the frequency of rare events as a weighted average of the frequencies of an event and a pool of events. The pool will draw upon, for example, events with similar precursors. The higher the degree of homogeneity of the pool, then the Empirical Bayes estimator will be more accurate. We propose and evaluate a new method using homogenisation factors under the assumption that events are generated from a Homogeneous Poisson Process. The homogenisation factors are scaling constants, which can be elicited through structured expert judgement and used to align the frequencies of different events, hence homogenising the pool. The estimation error relative to the homogeneity of the pool is examined theoretically indicating that reduced error is associated with larger pool homogeneity. The effects of misspecified expert assessments of the homogenisation factors are examined theoretically and through simulation experiments. Our results show that the proposed Empirical Bayes method using homogenisation factors is robust under different degrees of misspecification.",
keywords = "poisson processes, empirical bayes, pairwise comparison, PRA",
author = "John Quigley and Gavin Hardman and Tim Bedford and Lesley Walls",
year = "2011",
month = "6",
doi = "10.1016/j.ress.2010.12.007",
language = "English",
volume = "96",
pages = "687--695",
journal = "Reliability Engineering and System Safety",
issn = "0951-8320",
number = "6",

}

TY - JOUR

T1 - Merging expert and empirical data for rare event frequency estimation

T2 - Reliability Engineering and System Safety

AU - Quigley, John

AU - Hardman, Gavin

AU - Bedford, Tim

AU - Walls, Lesley

PY - 2011/6

Y1 - 2011/6

N2 - Empirical Bayes provides one approach to estimating the frequency of rare events as a weighted average of the frequencies of an event and a pool of events. The pool will draw upon, for example, events with similar precursors. The higher the degree of homogeneity of the pool, then the Empirical Bayes estimator will be more accurate. We propose and evaluate a new method using homogenisation factors under the assumption that events are generated from a Homogeneous Poisson Process. The homogenisation factors are scaling constants, which can be elicited through structured expert judgement and used to align the frequencies of different events, hence homogenising the pool. The estimation error relative to the homogeneity of the pool is examined theoretically indicating that reduced error is associated with larger pool homogeneity. The effects of misspecified expert assessments of the homogenisation factors are examined theoretically and through simulation experiments. Our results show that the proposed Empirical Bayes method using homogenisation factors is robust under different degrees of misspecification.

AB - Empirical Bayes provides one approach to estimating the frequency of rare events as a weighted average of the frequencies of an event and a pool of events. The pool will draw upon, for example, events with similar precursors. The higher the degree of homogeneity of the pool, then the Empirical Bayes estimator will be more accurate. We propose and evaluate a new method using homogenisation factors under the assumption that events are generated from a Homogeneous Poisson Process. The homogenisation factors are scaling constants, which can be elicited through structured expert judgement and used to align the frequencies of different events, hence homogenising the pool. The estimation error relative to the homogeneity of the pool is examined theoretically indicating that reduced error is associated with larger pool homogeneity. The effects of misspecified expert assessments of the homogenisation factors are examined theoretically and through simulation experiments. Our results show that the proposed Empirical Bayes method using homogenisation factors is robust under different degrees of misspecification.

KW - poisson processes

KW - empirical bayes

KW - pairwise comparison

KW - PRA

UR - http://www.journals.elsevier.com/reliability-engineering-and-system-safety

U2 - 10.1016/j.ress.2010.12.007

DO - 10.1016/j.ress.2010.12.007

M3 - Article

VL - 96

SP - 687

EP - 695

JO - Reliability Engineering and System Safety

JF - Reliability Engineering and System Safety

SN - 0951-8320

IS - 6

ER -