Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding

Autores
Caiafa, César Federico; Wang, Ziyao; Sole Casals, Jordi; Zhao, Qibin
Año de publicación
2021
Idioma
inglés
Tipo de recurso
documento de conferencia
Estado
versión publicada
Descripción
In this paper, the problem of training a classifier on a dataset with incomplete features is addressed. We assume that different subsets of features (random or structured) are available at each data instance. This situation typically occurs in the applications when not all the features are collected for every data sample. A new supervised learning method is developed to train a general classifier, such as a logistic regression or a deep neural network, using only a subset of features per sample, while assuming sparse representations of data vectors on an unknown dictionary. Sufficient conditions are identified, such that, if it is possible to train a classifier on incomplete observations so that their reconstructions are well separated by a hyperplane, then the same classifier also correctly separates the original (unobserved) data samples. Extensive simulation results on synthetic and well-known datasets are presented that validate our theoretical findings and demonstrate the effectiveness of the proposed method compared to traditional data imputation approaches and one state-of-the-art algorithm.
Fil: Caiafa, César Federico. Provincia de Buenos Aires. Gobernación. Comisión de Investigaciones Científicas. Instituto Argentino de Radioastronomía. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto Argentino de Radioastronomía; Argentina
Fil: Wang, Ziyao. South East University; China
Fil: Sole Casals, Jordi. University of Vic; España
Fil: Zhao, Qibin. Center for Advanced Intelligence Project; Japón
IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2021
New York
Estados Unidos
IEEE
Materia
Supervised learning
Missing data
Deep learning
Sparse Coding
Nivel de accesibilidad
acceso abierto
Condiciones de uso
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
Repositorio
CONICET Digital (CONICET)
Institución
Consejo Nacional de Investigaciones Científicas y Técnicas
OAI Identificador
oai:ri.conicet.gov.ar:11336/139273
Acceder
id CONICETDig_3df31e005616413a4a3a745347734c4c
oai_identifier_str oai:ri.conicet.gov.ar:11336/139273
network_acronym_str CONICETDig
repository_id_str 3498
network_name_str CONICET Digital (CONICET)
spelling Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse CodingCaiafa, César FedericoWang, ZiyaoSole Casals, JordiZhao, QibinSupervised learningMissing dataDeep learningSparse Codinghttps://purl.org/becyt/ford/1.2https://purl.org/becyt/ford/1In this paper, the problem of training a classifier on a dataset with incomplete features is addressed. We assume that different subsets of features (random or structured) are available at each data instance. This situation typically occurs in the applications when not all the features are collected for every data sample. A new supervised learning method is developed to train a general classifier, such as a logistic regression or a deep neural network, using only a subset of features per sample, while assuming sparse representations of data vectors on an unknown dictionary. Sufficient conditions are identified, such that, if it is possible to train a classifier on incomplete observations so that their reconstructions are well separated by a hyperplane, then the same classifier also correctly separates the original (unobserved) data samples. Extensive simulation results on synthetic and well-known datasets are presented that validate our theoretical findings and demonstrate the effectiveness of the proposed method compared to traditional data imputation approaches and one state-of-the-art algorithm.Fil: Caiafa, César Federico. Provincia de Buenos Aires. Gobernación. Comisión de Investigaciones Científicas. Instituto Argentino de Radioastronomía. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto Argentino de Radioastronomía; ArgentinaFil: Wang, Ziyao. South East University; ChinaFil: Sole Casals, Jordi. University of Vic; EspañaFil: Zhao, Qibin. Center for Advanced Intelligence Project; JapónIEEE Computer Society Conference on Computer Vision and Pattern Recognition 2021New YorkEstados UnidosIEEEIEEE2021info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObjectConferenciaBookhttp://purl.org/coar/resource_type/c_5794info:ar-repo/semantics/documentoDeConferenciaapplication/pdfapplication/pdfhttp://hdl.handle.net/11336/139273Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding; IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2021; New York; Estados Unidos; 2021; 1-11CONICET DigitalCONICETenginfo:eu-repo/semantics/altIdentifier/url/https://l2id.github.io/index.htmlinfo:eu-repo/semantics/altIdentifier/url/https://l2id.github.io/L2ID@CVPR2021_Accepted_paper_list.htmlinfo:eu-repo/semantics/altIdentifier/url/https://arxiv.org/abs/2011.14047Internacionalinfo:eu-repo/semantics/openAccesshttps://creativecommons.org/licenses/by-nc-sa/2.5/ar/reponame:CONICET Digital (CONICET)instname:Consejo Nacional de Investigaciones Científicas y Técnicas2025-09-29T10:02:01Zoai:ri.conicet.gov.ar:11336/139273instacron:CONICETInstitucionalhttp://ri.conicet.gov.ar/Organismo científico-tecnológicoNo correspondehttp://ri.conicet.gov.ar/oai/requestdasensio@conicet.gov.ar; lcarlino@conicet.gov.arArgentinaNo correspondeNo correspondeNo correspondeopendoar:34982025-09-29 10:02:01.992CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicasfalse
dc.title.none.fl_str_mv Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
title Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
spellingShingle Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
Caiafa, César Federico
Supervised learning
Missing data
Deep learning
Sparse Coding
title_short Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
title_full Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
title_fullStr Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
title_full_unstemmed Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
title_sort Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding
dc.creator.none.fl_str_mv Caiafa, César Federico
Wang, Ziyao
Sole Casals, Jordi
Zhao, Qibin
author Caiafa, César Federico
author_facet Caiafa, César Federico
Wang, Ziyao
Sole Casals, Jordi
Zhao, Qibin
author_role author
author2 Wang, Ziyao
Sole Casals, Jordi
Zhao, Qibin
author2_role author
author
author
dc.subject.none.fl_str_mv Supervised learning
Missing data
Deep learning
Sparse Coding
topic Supervised learning
Missing data
Deep learning
Sparse Coding
purl_subject.fl_str_mv https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
dc.description.none.fl_txt_mv In this paper, the problem of training a classifier on a dataset with incomplete features is addressed. We assume that different subsets of features (random or structured) are available at each data instance. This situation typically occurs in the applications when not all the features are collected for every data sample. A new supervised learning method is developed to train a general classifier, such as a logistic regression or a deep neural network, using only a subset of features per sample, while assuming sparse representations of data vectors on an unknown dictionary. Sufficient conditions are identified, such that, if it is possible to train a classifier on incomplete observations so that their reconstructions are well separated by a hyperplane, then the same classifier also correctly separates the original (unobserved) data samples. Extensive simulation results on synthetic and well-known datasets are presented that validate our theoretical findings and demonstrate the effectiveness of the proposed method compared to traditional data imputation approaches and one state-of-the-art algorithm.
Fil: Caiafa, César Federico. Provincia de Buenos Aires. Gobernación. Comisión de Investigaciones Científicas. Instituto Argentino de Radioastronomía. Consejo Nacional de Investigaciones Científicas y Técnicas. Centro Científico Tecnológico Conicet - La Plata. Instituto Argentino de Radioastronomía; Argentina
Fil: Wang, Ziyao. South East University; China
Fil: Sole Casals, Jordi. University of Vic; España
Fil: Zhao, Qibin. Center for Advanced Intelligence Project; Japón
IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2021
New York
Estados Unidos
IEEE
description In this paper, the problem of training a classifier on a dataset with incomplete features is addressed. We assume that different subsets of features (random or structured) are available at each data instance. This situation typically occurs in the applications when not all the features are collected for every data sample. A new supervised learning method is developed to train a general classifier, such as a logistic regression or a deep neural network, using only a subset of features per sample, while assuming sparse representations of data vectors on an unknown dictionary. Sufficient conditions are identified, such that, if it is possible to train a classifier on incomplete observations so that their reconstructions are well separated by a hyperplane, then the same classifier also correctly separates the original (unobserved) data samples. Extensive simulation results on synthetic and well-known datasets are presented that validate our theoretical findings and demonstrate the effectiveness of the proposed method compared to traditional data imputation approaches and one state-of-the-art algorithm.
publishDate 2021
dc.date.none.fl_str_mv 2021
dc.type.none.fl_str_mv info:eu-repo/semantics/publishedVersion
info:eu-repo/semantics/conferenceObject
Conferencia
Book
http://purl.org/coar/resource_type/c_5794
info:ar-repo/semantics/documentoDeConferencia
status_str publishedVersion
format conferenceObject
dc.identifier.none.fl_str_mv http://hdl.handle.net/11336/139273
Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding; IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2021; New York; Estados Unidos; 2021; 1-11
CONICET Digital
CONICET
url http://hdl.handle.net/11336/139273
identifier_str_mv Learning from Incomplete Features by Simultaneous Training of Neural Networks and Sparse Coding; IEEE Computer Society Conference on Computer Vision and Pattern Recognition 2021; New York; Estados Unidos; 2021; 1-11
CONICET Digital
CONICET
dc.language.none.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv info:eu-repo/semantics/altIdentifier/url/https://l2id.github.io/index.html
info:eu-repo/semantics/altIdentifier/url/https://l2id.github.io/L2ID@CVPR2021_Accepted_paper_list.html
info:eu-repo/semantics/altIdentifier/url/https://arxiv.org/abs/2011.14047
dc.rights.none.fl_str_mv info:eu-repo/semantics/openAccess
https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
eu_rights_str_mv openAccess
rights_invalid_str_mv https://creativecommons.org/licenses/by-nc-sa/2.5/ar/
dc.format.none.fl_str_mv application/pdf
application/pdf
dc.coverage.none.fl_str_mv Internacional
dc.publisher.none.fl_str_mv IEEE
publisher.none.fl_str_mv IEEE
dc.source.none.fl_str_mv reponame:CONICET Digital (CONICET)
instname:Consejo Nacional de Investigaciones Científicas y Técnicas
reponame_str CONICET Digital (CONICET)
collection CONICET Digital (CONICET)
instname_str Consejo Nacional de Investigaciones Científicas y Técnicas
repository.name.fl_str_mv CONICET Digital (CONICET) - Consejo Nacional de Investigaciones Científicas y Técnicas
repository.mail.fl_str_mv dasensio@conicet.gov.ar; lcarlino@conicet.gov.ar
_version_ 1844613819944402944
score 13.070432

Publicaciones similares