A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification
Feature selection is a widespread preprocessing step in the data mining field. One of its purposes is to reduce the number of original dataset features to improve a predictive model’s performance. Despite the benefits of feature selection for the classification task, to the best of our kn...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9536739/ |
| _version_ | 1811274347225022464 |
|---|---|
| author | Helen C. S. C. Lima Fernando E. B. Otero Luiz H. C. Merschmann Marcone J. F. Souza |
| author_facet | Helen C. S. C. Lima Fernando E. B. Otero Luiz H. C. Merschmann Marcone J. F. Souza |
| author_sort | Helen C. S. C. Lima |
| collection | DOAJ |
| description | Feature selection is a widespread preprocessing step in the data mining field. One of its purposes is to reduce the number of original dataset features to improve a predictive model’s performance. Despite the benefits of feature selection for the classification task, to the best of our knowledge, few studies in the literature address feature selection for the hierarchical classification context. This paper proposes a novel feature selection method based on the general variable neighborhood search metaheuristic, combining a filter and a wrapper step, wherein a global model hierarchical classifier evaluates feature subsets. We used twelve datasets from the proteins and images domains to perform computational experiments to validate the effect of the proposed algorithm on classification performance when using two global hierarchical classifiers proposed in the literature. Statistical tests showed that using our method for feature selection led to predictive performances that were consistently better than or equivalent to that obtained by using all features with the benefit of reducing the number of features needed, which justifies its efficiency for the hierarchical classification scenario. |
| first_indexed | 2024-04-12T23:17:25Z |
| format | Article |
| id | doaj.art-89ea2c2b2e894bb0b1b945ee7a76804a |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-12T23:17:25Z |
| publishDate | 2021-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-89ea2c2b2e894bb0b1b945ee7a76804a2022-12-22T03:12:38ZengIEEEIEEE Access2169-35362021-01-01912727812729210.1109/ACCESS.2021.31123969536739A Novel Hybrid Feature Selection Algorithm for Hierarchical ClassificationHelen C. S. C. Lima0https://orcid.org/0000-0002-2491-4750Fernando E. B. Otero1https://orcid.org/0000-0003-2172-297XLuiz H. C. Merschmann2https://orcid.org/0000-0002-9948-2673Marcone J. F. Souza3https://orcid.org/0000-0002-7141-357XDepartment of Computing, Federal University of Ouro Preto, Ouro Preto, BrazilSchool of Computing, University of Kent, Canterbury, U.K.Department of Applied Computing, Federal University of Lavras, Lavras, BrazilDepartment of Computing, Federal University of Ouro Preto, Ouro Preto, BrazilFeature selection is a widespread preprocessing step in the data mining field. One of its purposes is to reduce the number of original dataset features to improve a predictive model’s performance. Despite the benefits of feature selection for the classification task, to the best of our knowledge, few studies in the literature address feature selection for the hierarchical classification context. This paper proposes a novel feature selection method based on the general variable neighborhood search metaheuristic, combining a filter and a wrapper step, wherein a global model hierarchical classifier evaluates feature subsets. We used twelve datasets from the proteins and images domains to perform computational experiments to validate the effect of the proposed algorithm on classification performance when using two global hierarchical classifiers proposed in the literature. Statistical tests showed that using our method for feature selection led to predictive performances that were consistently better than or equivalent to that obtained by using all features with the benefit of reducing the number of features needed, which justifies its efficiency for the hierarchical classification scenario.https://ieeexplore.ieee.org/document/9536739/Feature selectionhierarchical single-label classificationvariable neighborhood searchfilterwrapper |
| spellingShingle | Helen C. S. C. Lima Fernando E. B. Otero Luiz H. C. Merschmann Marcone J. F. Souza A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification IEEE Access Feature selection hierarchical single-label classification variable neighborhood search filter wrapper |
| title | A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification |
| title_full | A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification |
| title_fullStr | A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification |
| title_full_unstemmed | A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification |
| title_short | A Novel Hybrid Feature Selection Algorithm for Hierarchical Classification |
| title_sort | novel hybrid feature selection algorithm for hierarchical classification |
| topic | Feature selection hierarchical single-label classification variable neighborhood search filter wrapper |
| url | https://ieeexplore.ieee.org/document/9536739/ |
| work_keys_str_mv | AT helencsclima anovelhybridfeatureselectionalgorithmforhierarchicalclassification AT fernandoebotero anovelhybridfeatureselectionalgorithmforhierarchicalclassification AT luizhcmerschmann anovelhybridfeatureselectionalgorithmforhierarchicalclassification AT marconejfsouza anovelhybridfeatureselectionalgorithmforhierarchicalclassification AT helencsclima novelhybridfeatureselectionalgorithmforhierarchicalclassification AT fernandoebotero novelhybridfeatureselectionalgorithmforhierarchicalclassification AT luizhcmerschmann novelhybridfeatureselectionalgorithmforhierarchicalclassification AT marconejfsouza novelhybridfeatureselectionalgorithmforhierarchicalclassification |