A Bayes linear Bayes method for estimation of correlated event rates

Research output: Contribution to journalArticle

4 Citations (Scopus)

Abstract

Typically, full Bayesian estimation of correlated event rates can be computationally challenging since estimators are intractable. When estimation of event rates represents one activity within a larger modeling process, there is an incentive to develop more efficient inference than provided by a full Bayesian model. We develop a new subjective inference method for correlated event rates based on a Bayes linear Bayes model under the assumption that events are generated from a homogeneous Poisson process. To reduce the elicitation burden we introduce homogenization factors to the model and, as an alternative to a subjective prior, an empirical method using the method of moments is developed. Inference under the new method is compared against estimates obtained under a full Bayesian model, which takes a multivariate gamma prior, where the predictive and posterior distributions are derived in terms of well-known functions. The mathematical properties of both models are presented. A simulation study shows that the Bayes linear Bayes inference method and the full Bayesian model provide equally reliable estimates. An illustrative example, motivated by a problem of estimating correlated event rates across different users in a simple supply chain, shows how ignoring the correlation leads to biased estimation of event rates.
LanguageEnglish
Pages2209–2224
Number of pages16
JournalRisk Analysis
Volume33
Issue number12
Early online date28 Mar 2013
DOIs
Publication statusPublished - Dec 2013

Fingerprint

Bayes Theorem
Motivation
Linear Models
Method of moments
Supply chains
Inference
Bayesian model
Predictive distribution
Posterior distribution
Simulation study
Supply chain
Homogenization
Incentives
Factors
Empirical methods
Burden
Process modeling
Estimator
Bayesian estimation
Poisson process

Keywords

  • reliability
  • correlated event rates
  • poisson process
  • bayes linear kinematics
  • empirical bayes
  • supply chain

Cite this

@article{58e3b12110504ea983d6554289d83dbf,
title = "A Bayes linear Bayes method for estimation of correlated event rates",
abstract = "Typically, full Bayesian estimation of correlated event rates can be computationally challenging since estimators are intractable. When estimation of event rates represents one activity within a larger modeling process, there is an incentive to develop more efficient inference than provided by a full Bayesian model. We develop a new subjective inference method for correlated event rates based on a Bayes linear Bayes model under the assumption that events are generated from a homogeneous Poisson process. To reduce the elicitation burden we introduce homogenization factors to the model and, as an alternative to a subjective prior, an empirical method using the method of moments is developed. Inference under the new method is compared against estimates obtained under a full Bayesian model, which takes a multivariate gamma prior, where the predictive and posterior distributions are derived in terms of well-known functions. The mathematical properties of both models are presented. A simulation study shows that the Bayes linear Bayes inference method and the full Bayesian model provide equally reliable estimates. An illustrative example, motivated by a problem of estimating correlated event rates across different users in a simple supply chain, shows how ignoring the correlation leads to biased estimation of event rates.",
keywords = "reliability, correlated event rates, poisson process, bayes linear kinematics, empirical bayes, supply chain",
author = "John Quigley and Kevin Wilson and Lesley Walls and Tim Bedford",
year = "2013",
month = "12",
doi = "10.1111/risa.12035",
language = "English",
volume = "33",
pages = "2209–2224",
journal = "Risk Analysis",
issn = "0272-4332",
number = "12",

}

A Bayes linear Bayes method for estimation of correlated event rates. / Quigley, John; Wilson, Kevin; Walls, Lesley; Bedford, Tim.

In: Risk Analysis, Vol. 33, No. 12, 12.2013, p. 2209–2224.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A Bayes linear Bayes method for estimation of correlated event rates

AU - Quigley, John

AU - Wilson, Kevin

AU - Walls, Lesley

AU - Bedford, Tim

PY - 2013/12

Y1 - 2013/12

N2 - Typically, full Bayesian estimation of correlated event rates can be computationally challenging since estimators are intractable. When estimation of event rates represents one activity within a larger modeling process, there is an incentive to develop more efficient inference than provided by a full Bayesian model. We develop a new subjective inference method for correlated event rates based on a Bayes linear Bayes model under the assumption that events are generated from a homogeneous Poisson process. To reduce the elicitation burden we introduce homogenization factors to the model and, as an alternative to a subjective prior, an empirical method using the method of moments is developed. Inference under the new method is compared against estimates obtained under a full Bayesian model, which takes a multivariate gamma prior, where the predictive and posterior distributions are derived in terms of well-known functions. The mathematical properties of both models are presented. A simulation study shows that the Bayes linear Bayes inference method and the full Bayesian model provide equally reliable estimates. An illustrative example, motivated by a problem of estimating correlated event rates across different users in a simple supply chain, shows how ignoring the correlation leads to biased estimation of event rates.

AB - Typically, full Bayesian estimation of correlated event rates can be computationally challenging since estimators are intractable. When estimation of event rates represents one activity within a larger modeling process, there is an incentive to develop more efficient inference than provided by a full Bayesian model. We develop a new subjective inference method for correlated event rates based on a Bayes linear Bayes model under the assumption that events are generated from a homogeneous Poisson process. To reduce the elicitation burden we introduce homogenization factors to the model and, as an alternative to a subjective prior, an empirical method using the method of moments is developed. Inference under the new method is compared against estimates obtained under a full Bayesian model, which takes a multivariate gamma prior, where the predictive and posterior distributions are derived in terms of well-known functions. The mathematical properties of both models are presented. A simulation study shows that the Bayes linear Bayes inference method and the full Bayesian model provide equally reliable estimates. An illustrative example, motivated by a problem of estimating correlated event rates across different users in a simple supply chain, shows how ignoring the correlation leads to biased estimation of event rates.

KW - reliability

KW - correlated event rates

KW - poisson process

KW - bayes linear kinematics

KW - empirical bayes

KW - supply chain

UR - http://onlinelibrary.wiley.com/journal/10.1111/%28ISSN%291539-6924

U2 - 10.1111/risa.12035

DO - 10.1111/risa.12035

M3 - Article

VL - 33

SP - 2209

EP - 2224

JO - Risk Analysis

T2 - Risk Analysis

JF - Risk Analysis

SN - 0272-4332

IS - 12

ER -