Saddlepoint approximation in exponential models with boundary points

Joan del Castillo; Anna López-Ratera

doi:https://doi.org/10.3150/bj/1151525132

Saddlepoint approximation in exponential models with boundary points

Joan del Castillo, Anna López-Ratera

Department of Mathematics

Research output: Contribution to journal › Article › Research › peer-review

6 Citations (Scopus)

Abstract

A saddlepoint approximation is developed for the likelihood ratio test statistic in non-regular exponential models. For the one-parameter case, it is proved that the saddlepoint correction has a relative error of order O(n-1), assuming that the canonical statistic has four finite moments. The results are applied in reliability theory and survival analysis, testing for exponentiality. Note that we are using the saddlepoint method for statistics without moment generating function. © 2006 ISI/BS.

Original language	English
Pages (from-to)	491-500
Journal	Bernoulli
Volume	12
DOIs	https://doi.org/10.3150/bj/1151525132
Publication status	Published - 1 Jun 2006

Keywords

Higher-order asymptotics
Likelihood ratio test
Reliability theory
Survival analysis models

Access to Document

https://doi.org/10.3150/bj/1151525132

Cite this

@article{b6f173b1f337411ea8b44c9c995d99b6,

title = "Saddlepoint approximation in exponential models with boundary points",

abstract = "A saddlepoint approximation is developed for the likelihood ratio test statistic in non-regular exponential models. For the one-parameter case, it is proved that the saddlepoint correction has a relative error of order O(n-1), assuming that the canonical statistic has four finite moments. The results are applied in reliability theory and survival analysis, testing for exponentiality. Note that we are using the saddlepoint method for statistics without moment generating function. {\textcopyright} 2006 ISI/BS.",

keywords = "Higher-order asymptotics, Likelihood ratio test, Reliability theory, Survival analysis models",

author = "{del Castillo}, Joan and Anna L{\'o}pez-Ratera",

year = "2006",

month = jun,

day = "1",

doi = "https://doi.org/10.3150/bj/1151525132",

language = "English",

volume = "12",

pages = "491--500",

journal = "Bernoulli",

issn = "1350-7265",

}

TY - JOUR

T1 - Saddlepoint approximation in exponential models with boundary points

AU - del Castillo, Joan

AU - López-Ratera, Anna

PY - 2006/6/1

Y1 - 2006/6/1

N2 - A saddlepoint approximation is developed for the likelihood ratio test statistic in non-regular exponential models. For the one-parameter case, it is proved that the saddlepoint correction has a relative error of order O(n-1), assuming that the canonical statistic has four finite moments. The results are applied in reliability theory and survival analysis, testing for exponentiality. Note that we are using the saddlepoint method for statistics without moment generating function. © 2006 ISI/BS.

AB - A saddlepoint approximation is developed for the likelihood ratio test statistic in non-regular exponential models. For the one-parameter case, it is proved that the saddlepoint correction has a relative error of order O(n-1), assuming that the canonical statistic has four finite moments. The results are applied in reliability theory and survival analysis, testing for exponentiality. Note that we are using the saddlepoint method for statistics without moment generating function. © 2006 ISI/BS.

KW - Higher-order asymptotics

KW - Likelihood ratio test

KW - Reliability theory

KW - Survival analysis models

U2 - https://doi.org/10.3150/bj/1151525132

DO - https://doi.org/10.3150/bj/1151525132

M3 - Article

SN - 1350-7265

VL - 12

SP - 491

EP - 500

JO - Bernoulli

JF - Bernoulli

ER -

Saddlepoint approximation in exponential models with boundary points

Abstract

Keywords

Access to Document

Fingerprint

Cite this