Balanced boosting with parallel perceptrons

Biblos-e Archivo/Manakin Repository

Show simple item record Cantador, Iván Dorronsoro, José R.
dc.contributor.other UAM. Departamento de Ingeniería Informática es_ES 2015-03-18T12:42:38Z 2015-03-18T12:42:38Z 2005
dc.identifier.citation Computational Intelligence and Bioinspired Systems: 8th International Work-Conference on Artificial Neural Networks, IWANN 2005, Vilanova i la Geltrú, Barcelona, Spain, June 8-10, 2005. Proceedings. Lecture Notes in Computer Science, Volumen 3512. Springer 2005. 208-216. en_US
dc.identifier.isbn 978-3-540-26208-4 (print) en_US
dc.identifier.isbn 978-3-540-32106-4 (online) en_US
dc.identifier.issn 0302-9743 (print) en_US
dc.identifier.issn 1611-3349 (online) en_US
dc.description The final publication is available at Springer via en_US
dc.description Proceedings of 8th International Work-Conference on Artificial Neural Networks, IWANN 2005, Vilanova i la Geltrú, Barcelona, Spain, June 8-10, 2005. en_US
dc.description.abstract Boosting constructs a weighted classifier out of possibly weak learners by successively concentrating on those patterns harder to classify. While giving excellent results in many problems, its performance can deteriorate in the presence of patterns with incorrect labels. In this work we shall use parallel perceptrons (PP), a novel approach to the classical committee machines, to detect whether a pattern’s label may not be correct and also whether it is redundant in the sense of being well represented in the training sample by many other similar patterns. Among other things, PP allow to naturally define margins for hidden unit activations, that we shall use to define the above pattern types. This pattern type classification allows a more nuanced approach to boosting. In particular, the procedure we shall propose, balanced boosting, uses it to modify boosting distribution updates. As we shall illustrate numerically, balanced boosting gives very good results on relatively hard classification problems, particularly in some that present a marked imbalance between class sizes. en_US
dc.description.sponsorship With partial support of Spain’s CICyT, TIC 01–572. en_US
dc.format.extent 9 pág. es_ES
dc.format.mimetype application/pdf en
dc.language.iso eng en
dc.publisher Springer Berlin Heidelberg
dc.relation.ispartof Lecture Notes in Computer Science en_US
dc.rights © Springer-Verlag Berlin Heidelberg 2005
dc.subject.other Computation by Abstract Devices en_US
dc.subject.other Algorithm Analysis and Problem Complexity en_US
dc.subject.other Image Processing and Computer Vision en_US
dc.subject.other Pattern Recognition en_US
dc.subject.other Evolutionary Biology en_US
dc.title Balanced boosting with parallel perceptrons en_US
dc.type conferenceObject en
dc.type bookPart en
dc.subject.eciencia Informática es_ES
dc.identifier.doi 10.1007/11494669_26
dc.identifier.publicationfirstpage 208
dc.identifier.publicationlastpage 216
dc.identifier.publicationvolume 3512
dc.relation.eventdate June 8-10, 2005 en_US
dc.relation.eventnumber 8
dc.relation.eventplace Vilanova i la Geltrú, Barcelona (Spain) en_US
dc.relation.eventtitle 8th International Work-Conference on Artificial Neural Networks, IWANN 2005 en_US
dc.type.version info:eu-repo/semantics/acceptedVersion en Aprendizaje Automático (ING EPS-001) es_ES
dc.rights.accessRights openAccess en
dc.authorUAM Cantador Gutiérrez, Iván (261086)
dc.authorUAM Dorronsoro Ibero, José Ramón (259712)

Files in this item


This item appears in the following Collection(s)

Show simple item record