- -

Mask selective regularization for restricted Boltzmann machines

RiuNet: Institutional repository of the Polithecnic University of Valencia

Share/Send to

Cited by

Statistics

  • Estadisticas de Uso

Mask selective regularization for restricted Boltzmann machines

Show simple item record

Files in this item

dc.contributor.author Mansanet Sandín, Jorge es_ES
dc.contributor.author Albiol Colomer, Alberto es_ES
dc.contributor.author Paredes Palacios, Roberto es_ES
dc.contributor.author Albiol Colomer, Antonio José es_ES
dc.date.accessioned 2016-05-03T08:00:55Z
dc.date.available 2016-05-03T08:00:55Z
dc.date.issued 2015-10-01
dc.identifier.issn 0925-2312
dc.identifier.uri http://hdl.handle.net/10251/63391
dc.description.abstract In the present work, we propose to deal with two important issues regarding to the RBM's learning capabilities. First, the topology of the input space, and second, the sparseness of the RBM obtained. One problem of RBMs is that they do not take advantage of the topology of the input space. In order to alleviate this lack, we propose to use a surrogate of the mutual information of the input representation space to build a set of binary masks. This approach is general and not only applicable to images, thus it can be extended to other layers in the standard layer-by-layer unsupervised learning. On the other hand, we propose a selective application of two different regularization terms, L-1 and L-2, in order to ensure the sparseness of the representation and the generalization capabilities. Additionally, another interesting capability of our approach is the adaptation of the topology of the network during the learning phase by means of selecting the best set of binary masks that fit the current weights configuration. The performance of these new ideas is assessed with a set of experiments on different well-known corpus. (C) 2015 Elsevier B.V. All rights reserved. es_ES
dc.description.sponsorship This work was financially supported by the Ministerio de Ciencia e Innovacion (Spain), Plan Nacional de I+D+i, Grant TEC2009-09146, and the FPI Grant BES-2010-032945. en_EN
dc.language Inglés es_ES
dc.publisher Elsevier es_ES
dc.relation.ispartof Neurocomputing es_ES
dc.rights Reserva de todos los derechos es_ES
dc.subject Restricted Boltzmann Machines es_ES
dc.subject Deep belief networks es_ES
dc.subject Regularization es_ES
dc.subject.classification LENGUAJES Y SISTEMAS INFORMATICOS es_ES
dc.subject.classification TEORIA DE LA SEÑAL Y COMUNICACIONES es_ES
dc.title Mask selective regularization for restricted Boltzmann machines es_ES
dc.type Artículo es_ES
dc.identifier.doi 10.1016/j.neucom.2015.03.026
dc.relation.projectID info:eu-repo/grantAgreement/MICINN//BES-2010-032945/ES/BES-2010-032945/ es_ES
dc.relation.projectID info:eu-repo/grantAgreement/MICINN//TEC2009-09146/ES/Nuevas Tecnicas Para Video Vigilancia Inteligente/ es_ES
dc.rights.accessRights Abierto es_ES
dc.contributor.affiliation Universitat Politècnica de València. Departamento de Comunicaciones - Departament de Comunicacions es_ES
dc.contributor.affiliation Universitat Politècnica de València. Departamento de Sistemas Informáticos y Computación - Departament de Sistemes Informàtics i Computació es_ES
dc.contributor.affiliation Universitat Politècnica de València. Instituto Universitario de Telecomunicación y Aplicaciones Multimedia - Institut Universitari de Telecomunicacions i Aplicacions Multimèdia es_ES
dc.description.bibliographicCitation Mansanet Sandín, J.; Albiol Colomer, A.; Paredes Palacios, R.; Albiol Colomer, AJ. (2015). Mask selective regularization for restricted Boltzmann machines. Neurocomputing. 165:375-383. https://doi.org/10.1016/j.neucom.2015.03.026 es_ES
dc.description.accrualMethod S es_ES
dc.relation.publisherversion http://dx.doi.org/10.1016/j.neucom.2015.03.026 es_ES
dc.description.upvformatpinicio 375 es_ES
dc.description.upvformatpfin 383 es_ES
dc.type.version info:eu-repo/semantics/publishedVersion es_ES
dc.description.volume 165 es_ES
dc.relation.senia 309029 es_ES


This item appears in the following Collection(s)

Show simple item record