Variable selection for 1D regression models

David J. Olive; Douglas M. Hawkins

doi:10.1198/004017004000000590

Variable selection for 1D regression models

David J. Olive, Douglas M. Hawkins

Statistics (Twin Cities)

Research output: Contribution to specialist publication › Article

14 Scopus citations

Abstract

variable selection, the search for j relevant predictor variables from a group of p candidates, is a standard problem in regression analysis. The class of ID regression models is a broad class that includes generalized linear models. We show that existing variable selection algorithms, originally meant for multiple linear regression and based on ordinary least squares and Mallows's Cp, can also be used for ID models. Graphical aids for variable selection are also provided.

Original language	English (US)
Pages	43-50
Number of pages	8
Volume	47
No	1
Specialist publication	Technometrics
DOIs	https://doi.org/10.1198/004017004000000590
State	Published - Feb 2005

Bibliographical note

Funding Information:
The authors thank to the editor, associate editor, and referees for a number of helpful suggestions for improvement in the article. Stan Young and Dennis Cook read an earlier version of the manuscript. This work was supported by National Science Foundation grants DMS-02-02922, DMS-03-06304, DMS-98-03622, and ACI 9619020.

Keywords

C
Cook's distance
Generalized linear model
Outlier
Regression graphics
Single index model

Publisher link

10.1198/004017004000000590

Find It

Find It at U of M Libraries

Cite this

@misc{72c867e42cc2433d80197437f2b8a66b,

title = "Variable selection for 1D regression models",

abstract = "variable selection, the search for j relevant predictor variables from a group of p candidates, is a standard problem in regression analysis. The class of ID regression models is a broad class that includes generalized linear models. We show that existing variable selection algorithms, originally meant for multiple linear regression and based on ordinary least squares and Mallows's Cp, can also be used for ID models. Graphical aids for variable selection are also provided.",

keywords = "C, Cook's distance, Generalized linear model, Outlier, Regression graphics, Single index model",

author = "Olive, {David J.} and Hawkins, {Douglas M.}",

note = "Funding Information: The authors thank to the editor, associate editor, and referees for a number of helpful suggestions for improvement in the article. Stan Young and Dennis Cook read an earlier version of the manuscript. This work was supported by National Science Foundation grants DMS-02-02922, DMS-03-06304, DMS-98-03622, and ACI 9619020.",

year = "2005",

month = feb,

doi = "10.1198/004017004000000590",

language = "English (US)",

volume = "47",

pages = "43--50",

journal = "Technometrics",

issn = "0040-1706",

publisher = "American Statistical Association",

}

TY - GEN

T1 - Variable selection for 1D regression models

AU - Olive, David J.

AU - Hawkins, Douglas M.

N1 - Funding Information: The authors thank to the editor, associate editor, and referees for a number of helpful suggestions for improvement in the article. Stan Young and Dennis Cook read an earlier version of the manuscript. This work was supported by National Science Foundation grants DMS-02-02922, DMS-03-06304, DMS-98-03622, and ACI 9619020.

PY - 2005/2

Y1 - 2005/2

N2 - variable selection, the search for j relevant predictor variables from a group of p candidates, is a standard problem in regression analysis. The class of ID regression models is a broad class that includes generalized linear models. We show that existing variable selection algorithms, originally meant for multiple linear regression and based on ordinary least squares and Mallows's Cp, can also be used for ID models. Graphical aids for variable selection are also provided.

AB - variable selection, the search for j relevant predictor variables from a group of p candidates, is a standard problem in regression analysis. The class of ID regression models is a broad class that includes generalized linear models. We show that existing variable selection algorithms, originally meant for multiple linear regression and based on ordinary least squares and Mallows's Cp, can also be used for ID models. Graphical aids for variable selection are also provided.

KW - C

KW - Cook's distance

KW - Generalized linear model

KW - Outlier

KW - Regression graphics

KW - Single index model

UR - http://www.scopus.com/inward/record.url?scp=13444269128&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=13444269128&partnerID=8YFLogxK

U2 - 10.1198/004017004000000590

DO - 10.1198/004017004000000590

M3 - Article

AN - SCOPUS:13444269128

SN - 0040-1706

VL - 47

SP - 43

EP - 50

JO - Technometrics

JF - Technometrics

ER -

Variable selection for 1D regression models

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this