Pair copula constructions for multivariate discrete data

Anastasios Nicholas Panagiotelis, Claudia Czado, Harry Joe

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Multivariate discrete response data can be found in diverse fields, including econometrics, finance, biometrics, and psychometrics. Our contribution, through this study, is to introduce a new class of models for multivariate discrete data based on pair copula constructions (PCCs) that has two major advantages. First, by deriving the conditions under which any multivariate discrete distribution can be decomposed as a PCC, we show that discrete PCCs attain highly flexible dependence structures. Second, the computational burden of evaluating the likelihood for an m-dimensional discrete PCC only grows quadratically with m. This compares favorably to existing models for which computing the likelihood either requires the evaluation of 2 m terms or slow numerical integration methods. We demonstrate the high quality of inference function for margins and maximum likelihood estimates, both under a simulated setting and for an application to a longitudinal discrete dataset on headache severity. This article has online supplementary material.
Original languageEnglish
Pages (from-to)1063 - 1072
Number of pages10
JournalJournal of the American Statistical Association
Volume107
Issue number499
DOIs
Publication statusPublished - 2012

Cite this

@article{e7ac0af0447d4cf0ae5e1f4d29c1da0c,
title = "Pair copula constructions for multivariate discrete data",
abstract = "Multivariate discrete response data can be found in diverse fields, including econometrics, finance, biometrics, and psychometrics. Our contribution, through this study, is to introduce a new class of models for multivariate discrete data based on pair copula constructions (PCCs) that has two major advantages. First, by deriving the conditions under which any multivariate discrete distribution can be decomposed as a PCC, we show that discrete PCCs attain highly flexible dependence structures. Second, the computational burden of evaluating the likelihood for an m-dimensional discrete PCC only grows quadratically with m. This compares favorably to existing models for which computing the likelihood either requires the evaluation of 2 m terms or slow numerical integration methods. We demonstrate the high quality of inference function for margins and maximum likelihood estimates, both under a simulated setting and for an application to a longitudinal discrete dataset on headache severity. This article has online supplementary material.",
author = "Panagiotelis, {Anastasios Nicholas} and Claudia Czado and Harry Joe",
year = "2012",
doi = "10.1080/01621459.2012.682850",
language = "English",
volume = "107",
pages = "1063 -- 1072",
journal = "Journal of the American Statistical Association",
issn = "0162-1459",
publisher = "Taylor & Francis",
number = "499",

}

Pair copula constructions for multivariate discrete data. / Panagiotelis, Anastasios Nicholas; Czado, Claudia; Joe, Harry.

In: Journal of the American Statistical Association, Vol. 107, No. 499, 2012, p. 1063 - 1072.

Research output: Contribution to journalArticleResearchpeer-review

TY - JOUR

T1 - Pair copula constructions for multivariate discrete data

AU - Panagiotelis, Anastasios Nicholas

AU - Czado, Claudia

AU - Joe, Harry

PY - 2012

Y1 - 2012

N2 - Multivariate discrete response data can be found in diverse fields, including econometrics, finance, biometrics, and psychometrics. Our contribution, through this study, is to introduce a new class of models for multivariate discrete data based on pair copula constructions (PCCs) that has two major advantages. First, by deriving the conditions under which any multivariate discrete distribution can be decomposed as a PCC, we show that discrete PCCs attain highly flexible dependence structures. Second, the computational burden of evaluating the likelihood for an m-dimensional discrete PCC only grows quadratically with m. This compares favorably to existing models for which computing the likelihood either requires the evaluation of 2 m terms or slow numerical integration methods. We demonstrate the high quality of inference function for margins and maximum likelihood estimates, both under a simulated setting and for an application to a longitudinal discrete dataset on headache severity. This article has online supplementary material.

AB - Multivariate discrete response data can be found in diverse fields, including econometrics, finance, biometrics, and psychometrics. Our contribution, through this study, is to introduce a new class of models for multivariate discrete data based on pair copula constructions (PCCs) that has two major advantages. First, by deriving the conditions under which any multivariate discrete distribution can be decomposed as a PCC, we show that discrete PCCs attain highly flexible dependence structures. Second, the computational burden of evaluating the likelihood for an m-dimensional discrete PCC only grows quadratically with m. This compares favorably to existing models for which computing the likelihood either requires the evaluation of 2 m terms or slow numerical integration methods. We demonstrate the high quality of inference function for margins and maximum likelihood estimates, both under a simulated setting and for an application to a longitudinal discrete dataset on headache severity. This article has online supplementary material.

U2 - 10.1080/01621459.2012.682850

DO - 10.1080/01621459.2012.682850

M3 - Article

VL - 107

SP - 1063

EP - 1072

JO - Journal of the American Statistical Association

JF - Journal of the American Statistical Association

SN - 0162-1459

IS - 499

ER -