Model selection when a key parameter is constrained to be in an interval

M. Z. Hossain, M. L. King

Research output: Contribution to journalArticle

Abstract

This article considers model selection procedures based on choosing the model with the largest maximized log-likelihood minus a penalty, when key parameters are restricted to be in a closed interval. Its main emphasis is how these penalties might be chosen in small samples to give good properties of the resultant procedure. We illustrate two model selection problems in the context of Box-Cox transformations and their application to the linear regression model. Simulation results for both problems indicate that the new procedure clearly dominates existing procedures in terms of having higher probabilities of correctly selecting the true model.

Original languageEnglish
Pages (from-to)1270-1280
Number of pages11
JournalCommunications in Statistics: Simulation and Computation
Volume37
Issue number7
DOIs
Publication statusPublished - Aug 2008

Fingerprint

Model Selection
Interval
Penalty
Box-Cox Transformation
Closed interval
Selection Procedures
Linear Regression Model
Small Sample
Likelihood
Linear regression
Model
Simulation
Context

Keywords

  • Box-Cox transformations
  • Controlled information criterion
  • Nuisance parameters
  • Parametric bootstrap
  • Regression model

ASJC Scopus subject areas

  • Statistics and Probability
  • Modelling and Simulation

Cite this

Model selection when a key parameter is constrained to be in an interval. / Hossain, M. Z.; King, M. L.

In: Communications in Statistics: Simulation and Computation, Vol. 37, No. 7, 08.2008, p. 1270-1280.

Research output: Contribution to journalArticle

@article{2dd67f3fd40a49a598914bd35e4994eb,
title = "Model selection when a key parameter is constrained to be in an interval",
abstract = "This article considers model selection procedures based on choosing the model with the largest maximized log-likelihood minus a penalty, when key parameters are restricted to be in a closed interval. Its main emphasis is how these penalties might be chosen in small samples to give good properties of the resultant procedure. We illustrate two model selection problems in the context of Box-Cox transformations and their application to the linear regression model. Simulation results for both problems indicate that the new procedure clearly dominates existing procedures in terms of having higher probabilities of correctly selecting the true model.",
keywords = "Box-Cox transformations, Controlled information criterion, Nuisance parameters, Parametric bootstrap, Regression model",
author = "Hossain, {M. Z.} and King, {M. L.}",
year = "2008",
month = "8",
doi = "10.1080/03610910802203303",
language = "English",
volume = "37",
pages = "1270--1280",
journal = "Communications in Statistics Part B: Simulation and Computation",
issn = "0361-0918",
publisher = "Taylor and Francis Ltd.",
number = "7",

}

TY - JOUR

T1 - Model selection when a key parameter is constrained to be in an interval

AU - Hossain, M. Z.

AU - King, M. L.

PY - 2008/8

Y1 - 2008/8

N2 - This article considers model selection procedures based on choosing the model with the largest maximized log-likelihood minus a penalty, when key parameters are restricted to be in a closed interval. Its main emphasis is how these penalties might be chosen in small samples to give good properties of the resultant procedure. We illustrate two model selection problems in the context of Box-Cox transformations and their application to the linear regression model. Simulation results for both problems indicate that the new procedure clearly dominates existing procedures in terms of having higher probabilities of correctly selecting the true model.

AB - This article considers model selection procedures based on choosing the model with the largest maximized log-likelihood minus a penalty, when key parameters are restricted to be in a closed interval. Its main emphasis is how these penalties might be chosen in small samples to give good properties of the resultant procedure. We illustrate two model selection problems in the context of Box-Cox transformations and their application to the linear regression model. Simulation results for both problems indicate that the new procedure clearly dominates existing procedures in terms of having higher probabilities of correctly selecting the true model.

KW - Box-Cox transformations

KW - Controlled information criterion

KW - Nuisance parameters

KW - Parametric bootstrap

KW - Regression model

UR - http://www.scopus.com/inward/record.url?scp=49349102763&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=49349102763&partnerID=8YFLogxK

U2 - 10.1080/03610910802203303

DO - 10.1080/03610910802203303

M3 - Article

VL - 37

SP - 1270

EP - 1280

JO - Communications in Statistics Part B: Simulation and Computation

JF - Communications in Statistics Part B: Simulation and Computation

SN - 0361-0918

IS - 7

ER -