To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning
Autor(a) principal: | |
---|---|
Data de Publicação: | 2015 |
Outros Autores: | , , , |
Tipo de documento: | Artigo de conferência |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UNESP |
Texto Completo: | http://dx.doi.org/10.1109/IJCNN.2015.7280644 http://hdl.handle.net/11449/168078 |
Resumo: | Many classification algorithms, such as Neural Networks and Support Vector Machines, have a range of hyper-parameters that may strongly affect the predictive performance of the models induced by them. Hence, it is recommended to define the values of these hyper-parameters using optimization techniques. While these techniques usually converge to a good set of values, they typically have a high computational cost, because many candidate sets of values are evaluated during the optimization process. It is often not clear whether this will result in parameter settings that are significantly better than the default settings. When training time is limited, it may help to know when these parameters should definitely be tuned. In this study, we use meta-learning to predict when optimization techniques are expected to lead to models whose predictive performance is better than those obtained by using default parameter settings. Hence, we can choose to employ optimization techniques only when they are expected to improve performance, thus reducing the overall computational cost. We evaluate these meta-learning techniques on more than one hundred data sets. The experimental results show that it is possible to accurately predict when optimization techniques should be used instead of default values suggested by some machine learning libraries. |
id |
UNSP_b71cf5fe2b634826a8629c34befe4708 |
---|---|
oai_identifier_str |
oai:repositorio.unesp.br:11449/168078 |
network_acronym_str |
UNSP |
network_name_str |
Repositório Institucional da UNESP |
repository_id_str |
2946 |
spelling |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learningComputational modelingNickelNiobiumOptimizationRadio frequencySupport vector machinesTrainingMany classification algorithms, such as Neural Networks and Support Vector Machines, have a range of hyper-parameters that may strongly affect the predictive performance of the models induced by them. Hence, it is recommended to define the values of these hyper-parameters using optimization techniques. While these techniques usually converge to a good set of values, they typically have a high computational cost, because many candidate sets of values are evaluated during the optimization process. It is often not clear whether this will result in parameter settings that are significantly better than the default settings. When training time is limited, it may help to know when these parameters should definitely be tuned. In this study, we use meta-learning to predict when optimization techniques are expected to lead to models whose predictive performance is better than those obtained by using default parameter settings. Hence, we can choose to employ optimization techniques only when they are expected to improve performance, thus reducing the overall computational cost. We evaluate these meta-learning techniques on more than one hundred data sets. The experimental results show that it is possible to accurately predict when optimization techniques should be used instead of default values suggested by some machine learning libraries.Universidade de São Paulo (USP)Universidade Estadual Paulista (UNESP)Eindhoven University of Technology (TU/e)Ludwig-Maximilians-University MunichUniversidade Estadual Paulista (UNESP)Universidade de São Paulo (USP)Universidade Estadual Paulista (Unesp)Eindhoven University of Technology (TU/e)Ludwig-Maximilians-University MunichMantovani, Rafael G.Rossi, André L. D. [UNESP]Vanschoren, JoaquinBischl, BerndCarvalho, André C.P.L.F.2018-12-11T16:39:39Z2018-12-11T16:39:39Z2015-09-28info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObjecthttp://dx.doi.org/10.1109/IJCNN.2015.7280644Proceedings of the International Joint Conference on Neural Networks, v. 2015-September.http://hdl.handle.net/11449/16807810.1109/IJCNN.2015.72806442-s2.0-84944265251Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengProceedings of the International Joint Conference on Neural Networksinfo:eu-repo/semantics/openAccess2021-10-23T21:47:01Zoai:repositorio.unesp.br:11449/168078Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462021-10-23T21:47:01Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false |
dc.title.none.fl_str_mv |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
title |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
spellingShingle |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning Mantovani, Rafael G. Computational modeling Nickel Niobium Optimization Radio frequency Support vector machines Training |
title_short |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
title_full |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
title_fullStr |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
title_full_unstemmed |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
title_sort |
To tune or not to tune: Recommending when to adjust SVM hyper-parameters via meta-learning |
author |
Mantovani, Rafael G. |
author_facet |
Mantovani, Rafael G. Rossi, André L. D. [UNESP] Vanschoren, Joaquin Bischl, Bernd Carvalho, André C.P.L.F. |
author_role |
author |
author2 |
Rossi, André L. D. [UNESP] Vanschoren, Joaquin Bischl, Bernd Carvalho, André C.P.L.F. |
author2_role |
author author author author |
dc.contributor.none.fl_str_mv |
Universidade de São Paulo (USP) Universidade Estadual Paulista (Unesp) Eindhoven University of Technology (TU/e) Ludwig-Maximilians-University Munich |
dc.contributor.author.fl_str_mv |
Mantovani, Rafael G. Rossi, André L. D. [UNESP] Vanschoren, Joaquin Bischl, Bernd Carvalho, André C.P.L.F. |
dc.subject.por.fl_str_mv |
Computational modeling Nickel Niobium Optimization Radio frequency Support vector machines Training |
topic |
Computational modeling Nickel Niobium Optimization Radio frequency Support vector machines Training |
description |
Many classification algorithms, such as Neural Networks and Support Vector Machines, have a range of hyper-parameters that may strongly affect the predictive performance of the models induced by them. Hence, it is recommended to define the values of these hyper-parameters using optimization techniques. While these techniques usually converge to a good set of values, they typically have a high computational cost, because many candidate sets of values are evaluated during the optimization process. It is often not clear whether this will result in parameter settings that are significantly better than the default settings. When training time is limited, it may help to know when these parameters should definitely be tuned. In this study, we use meta-learning to predict when optimization techniques are expected to lead to models whose predictive performance is better than those obtained by using default parameter settings. Hence, we can choose to employ optimization techniques only when they are expected to improve performance, thus reducing the overall computational cost. We evaluate these meta-learning techniques on more than one hundred data sets. The experimental results show that it is possible to accurately predict when optimization techniques should be used instead of default values suggested by some machine learning libraries. |
publishDate |
2015 |
dc.date.none.fl_str_mv |
2015-09-28 2018-12-11T16:39:39Z 2018-12-11T16:39:39Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/conferenceObject |
format |
conferenceObject |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://dx.doi.org/10.1109/IJCNN.2015.7280644 Proceedings of the International Joint Conference on Neural Networks, v. 2015-September. http://hdl.handle.net/11449/168078 10.1109/IJCNN.2015.7280644 2-s2.0-84944265251 |
url |
http://dx.doi.org/10.1109/IJCNN.2015.7280644 http://hdl.handle.net/11449/168078 |
identifier_str_mv |
Proceedings of the International Joint Conference on Neural Networks, v. 2015-September. 10.1109/IJCNN.2015.7280644 2-s2.0-84944265251 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
Proceedings of the International Joint Conference on Neural Networks |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.source.none.fl_str_mv |
Scopus 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_ |
1799965711746990080 |