Fine-tuning Deep Belief Networks using Harmony Search

Detalhes bibliográficos
Autor(a) principal: Papa, Joao Paulo [UNESP]
Data de Publicação: 2016
Outros Autores: Scheirer, Walter, Cox, David Daniel
Tipo de documento: Artigo
Idioma: eng
Título da fonte: Repositório Institucional da UNESP
Texto Completo: http://hdl.handle.net/11449/161620
Resumo: In this paper, we deal with the problem of Deep Belief Networks (DBNs) parameters fine-tuning by means of a fast meta-heuristic approach named Harmony Search (HS). Although such deep learning-based technique has been widely used in the last years, more detailed studies about how to set its parameters may not be observed in the literature. We have shown we can obtain more accurate results comparing HS against with several of its variants, a random search and two variants of the well-known Hyperopt library. The experimental results were carried out in two public datasets considering the task of binary image reconstruction, three DBN learning algorithms and three layers. (C) 2015 Elsevier B.V. All rights reserved.
id UNSP_efaf5da4e82851734144709ef7c9bc4f
oai_identifier_str oai:repositorio.unesp.br:11449/161620
network_acronym_str UNSP
network_name_str Repositório Institucional da UNESP
repository_id_str 2946
spelling Fine-tuning Deep Belief Networks using Harmony SearchRestricted Boltzmann MachinesDeep Belief NetworksHarmony SearchMeta-heuristicsIn this paper, we deal with the problem of Deep Belief Networks (DBNs) parameters fine-tuning by means of a fast meta-heuristic approach named Harmony Search (HS). Although such deep learning-based technique has been widely used in the last years, more detailed studies about how to set its parameters may not be observed in the literature. We have shown we can obtain more accurate results comparing HS against with several of its variants, a random search and two variants of the well-known Hyperopt library. The experimental results were carried out in two public datasets considering the task of binary image reconstruction, three DBN learning algorithms and three layers. (C) 2015 Elsevier B.V. All rights reserved.Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)UNESP Univ Estadual Paulista, Dept Comp, Bauru, BrazilHarvard Univ, Ctr Brain Sci, Cambridge, MA 02138 USAUNESP Univ Estadual Paulista, Dept Comp, Bauru, BrazilFAPESP: 2013/20387-7FAPESP: 2014/16250-9CNPq: 303182/2011-3CNPq: 470571/2013-6CNPq: 306166/2014-3Elsevier B.V.Universidade Estadual Paulista (Unesp)Harvard UnivPapa, Joao Paulo [UNESP]Scheirer, WalterCox, David Daniel2018-11-26T16:40:42Z2018-11-26T16:40:42Z2016-09-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/article875-885application/pdfApplied Soft Computing. Amsterdam: Elsevier Science Bv, v. 46, p. 875-885, 2016.1568-4946http://hdl.handle.net/11449/161620WOS:000377999900063WOS000377999900063.pdfWeb of Sciencereponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengApplied Soft Computing1,199info:eu-repo/semantics/openAccess2024-04-23T16:11:00Zoai:repositorio.unesp.br:11449/161620Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T20:50:13.963258Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false
dc.title.none.fl_str_mv Fine-tuning Deep Belief Networks using Harmony Search
title Fine-tuning Deep Belief Networks using Harmony Search
spellingShingle Fine-tuning Deep Belief Networks using Harmony Search
Papa, Joao Paulo [UNESP]
Restricted Boltzmann Machines
Deep Belief Networks
Harmony Search
Meta-heuristics
title_short Fine-tuning Deep Belief Networks using Harmony Search
title_full Fine-tuning Deep Belief Networks using Harmony Search
title_fullStr Fine-tuning Deep Belief Networks using Harmony Search
title_full_unstemmed Fine-tuning Deep Belief Networks using Harmony Search
title_sort Fine-tuning Deep Belief Networks using Harmony Search
author Papa, Joao Paulo [UNESP]
author_facet Papa, Joao Paulo [UNESP]
Scheirer, Walter
Cox, David Daniel
author_role author
author2 Scheirer, Walter
Cox, David Daniel
author2_role author
author
dc.contributor.none.fl_str_mv Universidade Estadual Paulista (Unesp)
Harvard Univ
dc.contributor.author.fl_str_mv Papa, Joao Paulo [UNESP]
Scheirer, Walter
Cox, David Daniel
dc.subject.por.fl_str_mv Restricted Boltzmann Machines
Deep Belief Networks
Harmony Search
Meta-heuristics
topic Restricted Boltzmann Machines
Deep Belief Networks
Harmony Search
Meta-heuristics
description In this paper, we deal with the problem of Deep Belief Networks (DBNs) parameters fine-tuning by means of a fast meta-heuristic approach named Harmony Search (HS). Although such deep learning-based technique has been widely used in the last years, more detailed studies about how to set its parameters may not be observed in the literature. We have shown we can obtain more accurate results comparing HS against with several of its variants, a random search and two variants of the well-known Hyperopt library. The experimental results were carried out in two public datasets considering the task of binary image reconstruction, three DBN learning algorithms and three layers. (C) 2015 Elsevier B.V. All rights reserved.
publishDate 2016
dc.date.none.fl_str_mv 2016-09-01
2018-11-26T16:40:42Z
2018-11-26T16:40:42Z
dc.type.status.fl_str_mv info:eu-repo/semantics/publishedVersion
dc.type.driver.fl_str_mv info:eu-repo/semantics/article
format article
status_str publishedVersion
dc.identifier.uri.fl_str_mv Applied Soft Computing. Amsterdam: Elsevier Science Bv, v. 46, p. 875-885, 2016.
1568-4946
http://hdl.handle.net/11449/161620
WOS:000377999900063
WOS000377999900063.pdf
identifier_str_mv Applied Soft Computing. Amsterdam: Elsevier Science Bv, v. 46, p. 875-885, 2016.
1568-4946
WOS:000377999900063
WOS000377999900063.pdf
url http://hdl.handle.net/11449/161620
dc.language.iso.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv Applied Soft Computing
1,199
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv 875-885
application/pdf
dc.publisher.none.fl_str_mv Elsevier B.V.
publisher.none.fl_str_mv Elsevier B.V.
dc.source.none.fl_str_mv Web of Science
reponame:Repositório Institucional da UNESP
instname:Universidade Estadual Paulista (UNESP)
instacron:UNESP
instname_str Universidade Estadual Paulista (UNESP)
instacron_str UNESP
institution UNESP
reponame_str Repositório Institucional da UNESP
collection Repositório Institucional da UNESP
repository.name.fl_str_mv Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)
repository.mail.fl_str_mv
_version_ 1808129256454094848