schliessen

Filtern

 

Bibliotheken

Network-Constrained Group Lasso for High-Dimensional Multinomial Classification with Application to Cancer Subtype Prediction

Classic multinomial logit model, commonly used in multiclass regression problem, is restricted to few predictors and does not take into account the relationship among variables. It has limited use for genomic data, where the number of genomic features far exceeds the sample size. Genomic features su... Full description

Journal Title: Cancer Informatics 2014, Vol.Suppl. 6(Suppl. 6), pp.25-33
Main Author: Tian, Xinyu
Other Authors: Wang, Xuefeng , Chen, Jun
Format: Electronic Article Electronic Article
Language: English
Quelle: © ProQuest LLC All rights reserved
ID: E-ISSN: 11769351
Zum Text:
SendSend as email Add to Book BagAdd to Book Bag
Staff View
recordid: proquest1695151377
title: Network-Constrained Group Lasso for High-Dimensional Multinomial Classification with Application to Cancer Subtype Prediction
format: Article
creator:
  • Tian, Xinyu
  • Wang, Xuefeng
  • Chen, Jun
ispartof: Cancer Informatics, 2014, Vol.Suppl. 6(Suppl. 6), pp.25-33
description: Classic multinomial logit model, commonly used in multiclass regression problem, is restricted to few predictors and does not take into account the relationship among variables. It has limited use for genomic data, where the number of genomic features far exceeds the sample size. Genomic features such as gene expressions are usually related by an underlying biological network. Efficient use of the network information is important to improve classification performance as well as the biological interpretability. We proposed a multinomial logit model that is capable of addressing both the high dimensionality of predictors and the underlying network information. Group lasso was used to induce model sparsity, and a network-constraint was imposed to induce the smoothness of the coefficients with respect to the underlying network structure. To deal with the non-smoothness of the objective function in optimization, we developed a proximal gradient algorithm for efficient computation. The proposed model was compared to models with no prior structure information in both simulations and a problem of cancer subtype prediction with real TCGA (the cancer genome atlas) gene expression data. The network-constrained mode outperformed the traditional ones in both cases.
language: eng
source: © ProQuest LLC All rights reserved
identifier: E-ISSN: 11769351
fulltext: fulltext_linktorsrc
issn:
  • 11769351
  • 1176-9351
url: Link


@attributes
ID813652337
RANK0.07
NO1
SEARCH_ENGINEprimo_central_multiple_fe
SEARCH_ENGINE_TYPEPrimo Central Search Engine
LOCALfalse
PrimoNMBib
record
control
sourcerecordid1695151377
sourceidproquest
recordidTN_proquest1695151377
sourcesystemOther
pqid1695151377
display
typearticle
titleNetwork-Constrained Group Lasso for High-Dimensional Multinomial Classification with Application to Cancer Subtype Prediction
creatorTian, Xinyu ; Wang, Xuefeng ; Chen, Jun
ispartofCancer Informatics, 2014, Vol.Suppl. 6(Suppl. 6), pp.25-33
identifierE-ISSN: 11769351
descriptionClassic multinomial logit model, commonly used in multiclass regression problem, is restricted to few predictors and does not take into account the relationship among variables. It has limited use for genomic data, where the number of genomic features far exceeds the sample size. Genomic features such as gene expressions are usually related by an underlying biological network. Efficient use of the network information is important to improve classification performance as well as the biological interpretability. We proposed a multinomial logit model that is capable of addressing both the high dimensionality of predictors and the underlying network information. Group lasso was used to induce model sparsity, and a network-constraint was imposed to induce the smoothness of the coefficients with respect to the underlying network structure. To deal with the non-smoothness of the objective function in optimization, we developed a proximal gradient algorithm for efficient computation. The proposed model was compared to models with no prior structure information in both simulations and a problem of cancer subtype prediction with real TCGA (the cancer genome atlas) gene expression data. The network-constrained mode outperformed the traditional ones in both cases.
languageeng
source
0© ProQuest LLC All rights reserved
1Advanced Technologies & Aerospace Database
2Biological Science Database
3ProQuest Computer Science Journals
4Australia & New Zealand Database
5Publicly Available Content Database
6Advanced Technologies Database with Aerospace
7Engineering Research Database
8Technology Research Database
9ProQuest Biological Science Collection
10ProQuest Engineering Collection
11ProQuest Environmental Science Collection
12ProQuest Advanced Technologies & Aerospace Collection
13ProQuest Natural Science Collection
14ProQuest Technology Collection
15ProQuest Computer Science Collection
16ProQuest SciTech Collection
17Computer and Information Systems Abstracts
18Agricultural & Environmental Science Database
19Materials Science & Engineering Database
20Natural Science Collection
21ProQuest Central (new)
22ProQuest Central K-12
23ProQuest Central Korea
24SciTech Premium Collection
25Technology Collection
26ProQuest Central Essentials
27ProQuest Central China
28ProQuest One Academic
29Environmental Science Index (ProQuest)
30Biological Science Index (ProQuest)
31Advanced Technologies & Aerospace Index (ProQuest)
32Engineering Index (ProQuest)
33Environmental Science Collection (ProQuest)
34Engineering Collection (ProQuest)
oafree_for_read
lds50peer_reviewed
links
openurl$$Topenurl_article
openurlfulltext$$Topenurlfull_article
linktorsrc$$Uhttp://search.proquest.com/docview/1695151377/?pq-origsite=primo$$EView_record_in_ProQuest_(subscribers_only)
search
creatorcontrib
0Tian, Xinyu
1Wang, Xuefeng
2Chen, Jun
titleNetwork-Constrained Group Lasso for High-Dimensional Multinomial Classification with Application to Cancer Subtype Prediction
descriptionClassic multinomial logit model, commonly used in multiclass regression problem, is restricted to few predictors and does not take into account the relationship among variables. It has limited use for genomic data, where the number of genomic features far exceeds the sample size. Genomic features such as gene expressions are usually related by an underlying biological network. Efficient use of the network information is important to improve classification performance as well as the biological interpretability. We proposed a multinomial logit model that is capable of addressing both the high dimensionality of predictors and the underlying network information. Group lasso was used to induce model sparsity, and a network-constraint was imposed to induce the smoothness of the coefficients with respect to the underlying network structure. To deal with the non-smoothness of the objective function in optimization, we developed a proximal gradient algorithm for efficient computation. The proposed model was compared to models with no prior structure information in both simulations and a problem of cancer subtype prediction with real TCGA (the cancer genome atlas) gene expression data. The network-constrained mode outperformed the traditional ones in both cases.
general
0English
1Sage Publications Ltd.
2Advanced Technologies & Aerospace Database
3Biological Science Database
4ProQuest Computer Science Journals
5Australia & New Zealand Database
6Publicly Available Content Database
7Advanced Technologies Database with Aerospace
8Engineering Research Database
9Technology Research Database
10ProQuest Biological Science Collection
11ProQuest Engineering Collection
12ProQuest Environmental Science Collection
13ProQuest Advanced Technologies & Aerospace Collection
14ProQuest Natural Science Collection
15ProQuest Technology Collection
16ProQuest Computer Science Collection
17ProQuest SciTech Collection
18Computer and Information Systems Abstracts
19Agricultural & Environmental Science Database
20Materials Science & Engineering Database
21Natural Science Collection
22ProQuest Central (new)
23ProQuest Central K-12
24ProQuest Central Korea
25SciTech Premium Collection
26Technology Collection
27ProQuest Central Essentials
28ProQuest Central China
29ProQuest One Academic
30Environmental Science Index (ProQuest)
31Biological Science Index (ProQuest)
32Advanced Technologies & Aerospace Index (ProQuest)
33Engineering Index (ProQuest)
34Environmental Science Collection (ProQuest)
35Engineering Collection (ProQuest)
sourceidproquest
recordidproquest1695151377
issn
011769351
11176-9351
rsrctypearticle
creationdate2014
addtitleCancer Informatics
searchscope
01007443
11007454
21007552
31007852
41007856
51007945
61008043
71008886
81009046
91009127
101009168
111009240
121009386
1310000002
1410000004
1510000005
1610000012
1710000013
1810000015
1910000022
2010000038
2110000041
2210000043
2310000045
2410000050
2510000053
2610000070
2710000120
2810000154
2910000195
3010000198
3110000199
3210000203
3310000209
3410000217
3510000233
3610000234
3710000238
3810000250
3910000253
4010000255
4110000256
4210000257
4310000258
4410000260
4510000265
4610000268
4710000281
4810000300
4910000348
5010000349
5110000350
5210000351
5310000353
5410000354
5510000356
5610000360
57proquest
scope
01007443
11007454
21007552
31007852
41007856
51007945
61008043
71008886
81009046
91009127
101009168
111009240
121009386
1310000002
1410000004
1510000005
1610000012
1710000013
1810000015
1910000022
2010000038
2110000041
2210000043
2310000045
2410000050
2510000053
2610000070
2710000120
2810000154
2910000195
3010000198
3110000199
3210000203
3310000209
3410000217
3510000233
3610000234
3710000238
3810000250
3910000253
4010000255
4110000256
4210000257
4310000258
4410000260
4510000265
4610000268
4710000281
4810000300
4910000348
5010000349
5110000350
5210000351
5310000353
5410000354
5510000356
5610000360
57proquest
lsr43
01007443false
11007454false
21007552false
31007852true
41007856true
51007945true
61008043true
71008886true
81009046true
91009127true
101009168true
111009240true
121009386true
1310000002false
1410000004false
1510000005false
1610000012false
1710000013false
1810000015false
1910000022false
2010000038true
2110000041false
2210000043false
2310000045true
2410000050true
2510000053true
2610000070true
2710000120true
2810000154false
2910000195false
3010000198false
3110000199false
3210000203false
3310000209false
3410000217false
3510000233true
3610000234false
3710000238true
3810000250false
3910000253true
4010000255true
4110000256true
4210000257true
4310000258true
4410000260true
4510000265true
4610000268true
4710000281true
4810000300true
4910000348true
5010000349false
5110000350false
5210000351false
5310000353false
5410000354false
5510000356false
5610000360true
startdate20140101
enddate20140101
citationpf 25 pt 33 vol suppl. 6 issue Suppl. 6
sort
titleNetwork-Constrained Group Lasso for High-Dimensional Multinomial Classification with Application to Cancer Subtype Prediction
authorTian, Xinyu ; Wang, Xuefeng ; Chen, Jun
creationdate20140101
lso0120140101
facets
frbrgroupid4226466927722872102
frbrtype6
languageeng
creationdate2014
collection
0Advanced Technologies & Aerospace Database
1Biological Science Database
2ProQuest Computer Science Journals
3Australia & New Zealand Database
4Publicly Available Content Database
5Advanced Technologies Database with Aerospace
6Engineering Research Database
7Technology Research Database
8ProQuest Biological Science Collection
9ProQuest Engineering Collection
10ProQuest Environmental Science Collection
11ProQuest Advanced Technologies & Aerospace Collection
12ProQuest Natural Science Collection
13ProQuest Technology Collection
14ProQuest Computer Science Collection
15ProQuest SciTech Collection
16Computer and Information Systems Abstracts
17Agricultural & Environmental Science Database
18Materials Science & Engineering Database
19Natural Science Collection
20ProQuest Central (new)
21ProQuest Central K-12
22ProQuest Central Korea
23SciTech Premium Collection
24Technology Collection
25ProQuest Central Essentials
26ProQuest Central China
27ProQuest One Academic
28Environmental Science Index (ProQuest)
29Biological Science Index (ProQuest)
30Advanced Technologies & Aerospace Index (ProQuest)
31Engineering Index (ProQuest)
32Environmental Science Collection (ProQuest)
33Engineering Collection (ProQuest)
prefilterarticles
rsrctypearticles
creatorcontrib
0Tian, Xinyu
1Wang, Xuefeng
2Chen, Jun
jtitleCancer Informatics
toplevelpeer_reviewed
delivery
delcategoryRemote Search Resource
fulltextfulltext_linktorsrc
addata
aulast
0Tian
1Wang
2Chen
aufirst
0Xinyu
1Xuefeng
2Jun
auinit1
0X.
1J.
au
0Tian, Xinyu
1Wang, Xuefeng
2Chen, Jun
atitleNetwork-Constrained Group Lasso for High-Dimensional Multinomial Classification with Application to Cancer Subtype Prediction
jtitleCancer Informatics
risdate20140101
volumesuppl. 6
issueSuppl. 6
spage25
epage33
pages25-33
eissn11769351
formatjournal
genrearticle
ristypeJOUR
abstractClassic multinomial logit model, commonly used in multiclass regression problem, is restricted to few predictors and does not take into account the relationship among variables. It has limited use for genomic data, where the number of genomic features far exceeds the sample size. Genomic features such as gene expressions are usually related by an underlying biological network. Efficient use of the network information is important to improve classification performance as well as the biological interpretability. We proposed a multinomial logit model that is capable of addressing both the high dimensionality of predictors and the underlying network information. Group lasso was used to induce model sparsity, and a network-constraint was imposed to induce the smoothness of the coefficients with respect to the underlying network structure. To deal with the non-smoothness of the objective function in optimization, we developed a proximal gradient algorithm for efficient computation. The proposed model was compared to models with no prior structure information in both simulations and a problem of cancer subtype prediction with real TCGA (the cancer genome atlas) gene expression data. The network-constrained mode outperformed the traditional ones in both cases.
copLondon
pubSage Publications Ltd.
urlhttp://search.proquest.com/docview/1695151377/
oafree_for_read
date2014-01-01