Exploiting diversity of margin-based classifiers

An experimental comparison among Support Vector Machines, AdaBoost and a recently proposed model for maximizing the margin with Feed-forward Neural Networks has been made on a real-world classification problem, namely Text Categorization. The results obtained when comparing their agreement on the pr...

Descripción completa

Detalles Bibliográficos
Autores: Romero Merino, Enrique|||0000-0003-2404-5716, Carreras Pérez, Xavier, Màrquez Villodre, Lluís|||0009-0009-0593-368X
Tipo de recurso: informe técnico
Fecha de publicación:2003
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/96843
Acceso en línea:https://hdl.handle.net/2117/96843
Access Level:acceso abierto
Palabra clave:Support Vector Machines
AdaBoost
Text categorization
Margin-based classifiers
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
id ES_765d4f32ddfad73a7e0bfaa4d571f37d
oai_identifier_str oai:upcommons.upc.edu:2117/96843
network_acronym_str ES
network_name_str España
repository_id_str
spelling Exploiting diversity of margin-based classifiersRomero Merino, Enrique|||0000-0003-2404-5716Carreras Pérez, XavierMàrquez Villodre, Lluís|||0009-0009-0593-368XSupport Vector MachinesAdaBoostText categorizationMargin-based classifiersÀrees temàtiques de la UPC::Informàtica::Intel·ligència artificialAn experimental comparison among Support Vector Machines, AdaBoost and a recently proposed model for maximizing the margin with Feed-forward Neural Networks has been made on a real-world classification problem, namely Text Categorization. The results obtained when comparing their agreement on the predictions show that similar performance does not imply similar predictions, suggesting that different models can be combined to obtain better performance. As a consequence of the study, we derived a very simple confidence measure of the prediction of the tested margin-based classifiers. This measure is based on the margin curve. The combination of margin-based classifiers with this confidence measure lead to a marked improvement on the performance of the system, when combined with several well-known combination schemes.20032003-12-0120162016-11-18reporthttp://purl.org/coar/resource_type/c_93fcVoRhttp://purl.org/coar/version/c_970fb48d4fbd8a85info:eu-repo/semantics/reportapplication/postscripthttps://hdl.handle.net/2117/96843reponame:UPCommons. Portal del coneixement obert de la UPCinstname:Universitat Politècnica de Catalunya (UPC)Inglésengopen accesshttp://purl.org/coar/access_right/c_abf2info:eu-repo/semantics/openAccessoai:upcommons.upc.edu:2117/968432026-05-27T15:37:01Z
dc.title.none.fl_str_mv Exploiting diversity of margin-based classifiers
title Exploiting diversity of margin-based classifiers
spellingShingle Exploiting diversity of margin-based classifiers
Romero Merino, Enrique|||0000-0003-2404-5716
Support Vector Machines
AdaBoost
Text categorization
Margin-based classifiers
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
title_short Exploiting diversity of margin-based classifiers
title_full Exploiting diversity of margin-based classifiers
title_fullStr Exploiting diversity of margin-based classifiers
title_full_unstemmed Exploiting diversity of margin-based classifiers
title_sort Exploiting diversity of margin-based classifiers
dc.creator.none.fl_str_mv Romero Merino, Enrique|||0000-0003-2404-5716
Carreras Pérez, Xavier
Màrquez Villodre, Lluís|||0009-0009-0593-368X
author Romero Merino, Enrique|||0000-0003-2404-5716
author_facet Romero Merino, Enrique|||0000-0003-2404-5716
Carreras Pérez, Xavier
Màrquez Villodre, Lluís|||0009-0009-0593-368X
author_role author
author2 Carreras Pérez, Xavier
Màrquez Villodre, Lluís|||0009-0009-0593-368X
author2_role author
author
dc.subject.none.fl_str_mv Support Vector Machines
AdaBoost
Text categorization
Margin-based classifiers
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
topic Support Vector Machines
AdaBoost
Text categorization
Margin-based classifiers
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
description An experimental comparison among Support Vector Machines, AdaBoost and a recently proposed model for maximizing the margin with Feed-forward Neural Networks has been made on a real-world classification problem, namely Text Categorization. The results obtained when comparing their agreement on the predictions show that similar performance does not imply similar predictions, suggesting that different models can be combined to obtain better performance. As a consequence of the study, we derived a very simple confidence measure of the prediction of the tested margin-based classifiers. This measure is based on the margin curve. The combination of margin-based classifiers with this confidence measure lead to a marked improvement on the performance of the system, when combined with several well-known combination schemes.
publishDate 2003
dc.date.none.fl_str_mv 2003
2003-12-01
2016
2016-11-18
dc.type.none.fl_str_mv report
http://purl.org/coar/resource_type/c_93fc
VoR
http://purl.org/coar/version/c_970fb48d4fbd8a85
dc.type.openaire.fl_str_mv info:eu-repo/semantics/report
format report
dc.identifier.none.fl_str_mv https://hdl.handle.net/2117/96843
url https://hdl.handle.net/2117/96843
dc.language.none.fl_str_mv Inglés
eng
language_invalid_str_mv Inglés
language eng
dc.rights.none.fl_str_mv open access
http://purl.org/coar/access_right/c_abf2
dc.rights.openaire.fl_str_mv info:eu-repo/semantics/openAccess
rights_invalid_str_mv open access
http://purl.org/coar/access_right/c_abf2
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/postscript
dc.source.none.fl_str_mv reponame:UPCommons. Portal del coneixement obert de la UPC
instname:Universitat Politècnica de Catalunya (UPC)
instname_str Universitat Politècnica de Catalunya (UPC)
reponame_str UPCommons. Portal del coneixement obert de la UPC
collection UPCommons. Portal del coneixement obert de la UPC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869411045744836608
score 15,300719

Similares en LA Referencia