schliessen

Filtern

 

Bibliotheken

ON BIC'S SELECTION CONSISTENCY FOR DISCRIMINANT ANALYSIS

Linear and quadratic discriminant analysis are two useful classification methods for which the problem of variable selection is of fundamental importance. To this end, a BIC-type selection criterion (Schwarz (1978)) was recently studied by Raftery and Dean (2006). Despite its usefulness, the BIC's s... Full description

Journal Title: Statistica Sinica 1 April 2011, Vol.21(2), pp.731-740
Main Author: Zhang, Qiong
Other Authors: Wang, Hansheng
Format: Electronic Article Electronic Article
Language: English
Subjects:
ID: ISSN: 10170405 ; E-ISSN: 19968507
Link: https://www.jstor.org/stable/24309538
Zum Text:
SendSend as email Add to Book BagAdd to Book Bag
Staff View
recordid: jstor_archive_3324309538
title: ON BIC'S SELECTION CONSISTENCY FOR DISCRIMINANT ANALYSIS
format: Article
creator:
  • Zhang, Qiong
  • Wang, Hansheng
subjects:
  • Applied sciences -- Research methods -- Modeling
  • Mathematics -- Applied mathematics -- Statistics
  • Information science -- Data products -- Datasets
  • Applied sciences -- Engineering -- Aerospace engineering
  • Mathematics -- Applied mathematics -- Statistics
  • Mathematics -- Applied mathematics -- Statistics
  • Mathematics -- Applied mathematics -- Statistics
  • Mathematics -- Applied mathematics -- Statistics
  • Applied sciences -- Engineering -- Aerospace engineering
ispartof: Statistica Sinica, 1 April 2011, Vol.21(2), pp.731-740
description: Linear and quadratic discriminant analysis are two useful classification methods for which the problem of variable selection is of fundamental importance. To this end, a BIC-type selection criterion (Schwarz (1978)) was recently studied by Raftery and Dean (2006). Despite its usefulness, the BIC's selection consistency (Shao (1997)) was not investigated. To fill this gap, we show that BIC, in conjunction with a backward elimination procedure, is indeed selection consistent. To confirm our asymptotic theory, a number of numerical studies are presented.
language: eng
source:
identifier: ISSN: 10170405 ; E-ISSN: 19968507
fulltext: fulltext
issn:
  • 1017-0405
  • 10170405
  • 1996-8507
  • 19968507
url: Link


@attributes
ID1789845054
RANK0.07
NO1
SEARCH_ENGINEprimo_central_multiple_fe
SEARCH_ENGINE_TYPEPrimo Central Search Engine
LOCALfalse
PrimoNMBib
record
control
sourcerecordid24309538
sourceidjstor_archive_33
recordidTN_jstor_archive_3324309538
sourceformatXML
sourcesystemPC
display
typearticle
titleON BIC'S SELECTION CONSISTENCY FOR DISCRIMINANT ANALYSIS
creatorZhang, Qiong ; Wang, Hansheng
ispartofStatistica Sinica, 1 April 2011, Vol.21(2), pp.731-740
identifierISSN: 10170405 ; E-ISSN: 19968507
subjectApplied sciences -- Research methods -- Modeling ; Mathematics -- Applied mathematics -- Statistics ; Information science -- Data products -- Datasets ; Applied sciences -- Engineering -- Aerospace engineering ; Mathematics -- Applied mathematics -- Statistics ; Mathematics -- Applied mathematics -- Statistics ; Mathematics -- Applied mathematics -- Statistics ; Mathematics -- Applied mathematics -- Statistics ; Applied sciences -- Engineering -- Aerospace engineering
descriptionLinear and quadratic discriminant analysis are two useful classification methods for which the problem of variable selection is of fundamental importance. To this end, a BIC-type selection criterion (Schwarz (1978)) was recently studied by Raftery and Dean (2006). Despite its usefulness, the BIC's selection consistency (Shao (1997)) was not investigated. To fill this gap, we show that BIC, in conjunction with a backward elimination procedure, is indeed selection consistent. To confirm our asymptotic theory, a number of numerical studies are presented.
languageeng
source
version2
lds50peer_reviewed
links
openurl$$Topenurl_article
backlink$$Uhttps://www.jstor.org/stable/24309538$$EView_this_record_in_JSTOR
openurlfulltext$$Topenurlfull_article
search
creatorcontrib
0Zhang, Qiong
1Wang, Hansheng
titleON BIC'S SELECTION CONSISTENCY FOR DISCRIMINANT ANALYSIS
descriptionLinear and quadratic discriminant analysis are two useful classification methods for which the problem of variable selection is of fundamental importance. To this end, a BIC-type selection criterion (Schwarz (1978)) was recently studied by Raftery and Dean (2006). Despite its usefulness, the BIC's selection consistency (Shao (1997)) was not investigated. To fill this gap, we show that BIC, in conjunction with a backward elimination procedure, is indeed selection consistent. To confirm our asymptotic theory, a number of numerical studies are presented.
subject
0Applied sciences -- Research methods -- Modeling
1Mathematics -- Applied mathematics -- Statistics
2Information science -- Data products -- Datasets
3Applied sciences -- Engineering -- Aerospace engineering
general
0English
1Institute of Statistical Science, Academia Sinica and International Chinese Statistical Association
2Archival Journals (JSTOR)
3JSTOR Archival Journals
sourceidjstor_archive_33
recordidjstor_archive_3324309538
issn
01017-0405
110170405
21996-8507
319968507
rsrctypearticle
creationdate2011
recordtypearticle
addtitleStatistica Sinica
searchscope
0jstor
1jstor_asc15
scope
0jstor
1jstor_asc15
citationpf 731
startdate20110401
enddate20110401
lsr30VSR-Enriched:[doi]
sort
titleON BIC'S SELECTION CONSISTENCY FOR DISCRIMINANT ANALYSIS
authorZhang, Qiong ; Wang, Hansheng
creationdate20110401
lso0120110401
facets
frbrgroupid8257226389188878323
frbrtype5
languageeng
creationdate2011
topic
0Applied sciences -- Research methods -- Modeling
1Mathematics -- Applied mathematics -- Statistics
2Information science -- Data products -- Datasets
3Applied sciences -- Engineering -- Aerospace engineering
collectionJSTOR Archival Journals
prefilterarticles
rsrctypearticles
creatorcontrib
0Zhang, Qiong
1Wang, Hansheng
jtitleStatistica Sinica
toplevelpeer_reviewed
delivery
delcategoryRemote Search Resource
fulltextfulltext
addata
aulast
0Zhang
1Wang
aufirst
0Qiong
1Hansheng
au
0Zhang, Qiong
1Wang, Hansheng
atitleON BIC'S SELECTION CONSISTENCY FOR DISCRIMINANT ANALYSIS
jtitleStatistica Sinica
risdate20110401
volume21
issue2
spage731
epage740
pages731-740
issn10170405
eissn19968507
formatjournal
genrearticle
ristypeJOUR
abstractLinear and quadratic discriminant analysis are two useful classification methods for which the problem of variable selection is of fundamental importance. To this end, a BIC-type selection criterion (Schwarz (1978)) was recently studied by Raftery and Dean (2006). Despite its usefulness, the BIC's selection consistency (Shao (1997)) was not investigated. To fill this gap, we show that BIC, in conjunction with a backward elimination procedure, is indeed selection consistent. To confirm our asymptotic theory, a number of numerical studies are presented.
pubInstitute of Statistical Science, Academia Sinica and International Chinese Statistical Association
doi10.5705/ss.2011.032a
date2011-04-01