Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms

Detalhes bibliográficos
Autor(a) principal: Zhao, J.
Data de Publicação: 2016
Outros Autores: Basto-Fernandes, V., Jiao, L., Yevseyeva, I., Maulana, A., Li, R., Bäck, T., Tang, K., Emmerichd, M. T. M.
Tipo de documento: Artigo
Idioma: eng
Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
Texto Completo: http://hdl.handle.net/10071/12775
Resumo: The receiver operating characteristic (ROC) and detection error tradeoff (DET) curves are frequently used in the machine learning community to analyze the performance of binary classifiers. Recently, the convex-hull-based multiobjective genetic programming algorithm was proposed and successfully applied to maximize the convex hull area for binary classification problems by minimizing false positive rate and maximizing true positive rate at the same time using indicator-based evolutionary algorithms. The area under the ROC curve was used for the performance assessment and to guide the search. Here we extend this research and propose two major advancements: Firstly we formulate the algorithm in detection error tradeoff space, minimizing false positives and false negatives, with the advantage that misclassification cost tradeoff can be assessed directly. Secondly, we add complexity as an objective function, which gives rise to a 3D objective space (as opposed to a 2D previous ROC space). A domain specific performance indicator for 3D Pareto front approximations, the volume above DET surface, is introduced, and used to guide the indicator -based evolutionary algorithm to find optimal approximation sets. We assess the performance of the new algorithm on designed theoretical problems with different geometries of Pareto fronts and DET surfaces, and two application-oriented benchmarks: (1) Designing spam filters with low numbers of false rejects, false accepts, and low computational cost using rule ensembles, and (2) finding sparse neural networks for binary classification of test data from the UCI machine learning benchmark. The results show a high performance of the new algorithm as compared to conventional methods for multicriteria optimization.
id RCAP_c826218e9562a62da46acecc2de8fbd7
oai_identifier_str oai:repositorio.iscte-iul.pt:10071/12775
network_acronym_str RCAP
network_name_str Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
repository_id_str 7160
spelling Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithmsConvex hullClassificationEvolutionary multiobjective optimizationParsimonyROC analysisAnti-spam filtersThe receiver operating characteristic (ROC) and detection error tradeoff (DET) curves are frequently used in the machine learning community to analyze the performance of binary classifiers. Recently, the convex-hull-based multiobjective genetic programming algorithm was proposed and successfully applied to maximize the convex hull area for binary classification problems by minimizing false positive rate and maximizing true positive rate at the same time using indicator-based evolutionary algorithms. The area under the ROC curve was used for the performance assessment and to guide the search. Here we extend this research and propose two major advancements: Firstly we formulate the algorithm in detection error tradeoff space, minimizing false positives and false negatives, with the advantage that misclassification cost tradeoff can be assessed directly. Secondly, we add complexity as an objective function, which gives rise to a 3D objective space (as opposed to a 2D previous ROC space). A domain specific performance indicator for 3D Pareto front approximations, the volume above DET surface, is introduced, and used to guide the indicator -based evolutionary algorithm to find optimal approximation sets. We assess the performance of the new algorithm on designed theoretical problems with different geometries of Pareto fronts and DET surfaces, and two application-oriented benchmarks: (1) Designing spam filters with low numbers of false rejects, false accepts, and low computational cost using rule ensembles, and (2) finding sparse neural networks for binary classification of test data from the UCI machine learning benchmark. The results show a high performance of the new algorithm as compared to conventional methods for multicriteria optimization.Elsevier2017-04-05T15:09:17Z2016-01-01T00:00:00Z20162019-04-12T11:37:08Zinfo:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/articleapplication/pdfhttp://hdl.handle.net/10071/12775eng0020-025510.1016/j.ins.2016.05.026Zhao, J.Basto-Fernandes, V.Jiao, L.Yevseyeva, I.Maulana, A.Li, R.Bäck, T.Tang, K.Emmerichd, M. T. M.info:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2023-11-09T17:51:29Zoai:repositorio.iscte-iul.pt:10071/12775Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-19T22:25:30.042651Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse
dc.title.none.fl_str_mv Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
title Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
spellingShingle Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
Zhao, J.
Convex hull
Classification
Evolutionary multiobjective optimization
Parsimony
ROC analysis
Anti-spam filters
title_short Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
title_full Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
title_fullStr Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
title_full_unstemmed Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
title_sort Multiobjective optimization of classifiers by means of 3-D convex Hull based evolutionary algorithms
author Zhao, J.
author_facet Zhao, J.
Basto-Fernandes, V.
Jiao, L.
Yevseyeva, I.
Maulana, A.
Li, R.
Bäck, T.
Tang, K.
Emmerichd, M. T. M.
author_role author
author2 Basto-Fernandes, V.
Jiao, L.
Yevseyeva, I.
Maulana, A.
Li, R.
Bäck, T.
Tang, K.
Emmerichd, M. T. M.
author2_role author
author
author
author
author
author
author
author
dc.contributor.author.fl_str_mv Zhao, J.
Basto-Fernandes, V.
Jiao, L.
Yevseyeva, I.
Maulana, A.
Li, R.
Bäck, T.
Tang, K.
Emmerichd, M. T. M.
dc.subject.por.fl_str_mv Convex hull
Classification
Evolutionary multiobjective optimization
Parsimony
ROC analysis
Anti-spam filters
topic Convex hull
Classification
Evolutionary multiobjective optimization
Parsimony
ROC analysis
Anti-spam filters
description The receiver operating characteristic (ROC) and detection error tradeoff (DET) curves are frequently used in the machine learning community to analyze the performance of binary classifiers. Recently, the convex-hull-based multiobjective genetic programming algorithm was proposed and successfully applied to maximize the convex hull area for binary classification problems by minimizing false positive rate and maximizing true positive rate at the same time using indicator-based evolutionary algorithms. The area under the ROC curve was used for the performance assessment and to guide the search. Here we extend this research and propose two major advancements: Firstly we formulate the algorithm in detection error tradeoff space, minimizing false positives and false negatives, with the advantage that misclassification cost tradeoff can be assessed directly. Secondly, we add complexity as an objective function, which gives rise to a 3D objective space (as opposed to a 2D previous ROC space). A domain specific performance indicator for 3D Pareto front approximations, the volume above DET surface, is introduced, and used to guide the indicator -based evolutionary algorithm to find optimal approximation sets. We assess the performance of the new algorithm on designed theoretical problems with different geometries of Pareto fronts and DET surfaces, and two application-oriented benchmarks: (1) Designing spam filters with low numbers of false rejects, false accepts, and low computational cost using rule ensembles, and (2) finding sparse neural networks for binary classification of test data from the UCI machine learning benchmark. The results show a high performance of the new algorithm as compared to conventional methods for multicriteria optimization.
publishDate 2016
dc.date.none.fl_str_mv 2016-01-01T00:00:00Z
2016
2017-04-05T15:09:17Z
2019-04-12T11:37:08Z
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 http://hdl.handle.net/10071/12775
url http://hdl.handle.net/10071/12775
dc.language.iso.fl_str_mv eng
language eng
dc.relation.none.fl_str_mv 0020-0255
10.1016/j.ins.2016.05.026
dc.rights.driver.fl_str_mv info:eu-repo/semantics/openAccess
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação
instacron:RCAAP
instname_str Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação
instacron_str RCAAP
institution RCAAP
reponame_str Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
collection Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
repository.name.fl_str_mv Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação
repository.mail.fl_str_mv
_version_ 1799134818252881920