Feature Selection via Pareto Multi-objective Genetic Algorithms

Feature Selection via Pareto Multi-objective Genetic Algorithms

Author Spolaor, Newton Google Scholar
Lorena, Ana Carolina Autor UNIFESP Google Scholar
Lee, Huei Diana Google Scholar
Abstract Feature selection, an important combinatorial optimization problem in data mining, aims to find a reduced subset of features of high quality in a dataset. Different categories of importance measures can be used to estimate the quality of a feature subset. Since each measure provides a distinct perspective of data and of which are their important features, in this article we investigate the simultaneous optimization of importance measures from different categories using multi-objective genetic algorithms grounded in the Pareto theory. An extensive experimental evaluation of the proposed method is presented, including an analysis of the performance of predictive models built using the selected subsets of features. The results show the competitiveness of the method in comparison with six feature selection algorithms. As an additional contribution, we conducted a pioneer, rigorous, and replicable systematic review on related work. As a result, a summary of 93 related papers strengthens features of our method.
xmlui.dri2xhtml.METS-1.0.item-coverage Philadelphia
Language English
Sponsor Brazilian National Council for Scientific and Technological Development (CNPq)
Sao Paulo Research Foundation (FAPESP)
Coordination for the Improvement of Higher Education Personnel (CAPES)
Federal University of ABC
Grant number CNPq: 482222/2013-1
CNPq: 308232/2011-9
FAPESP: 2012/22608-8
FAPESP: 2009/12963-2
Date 2017
Published in Applied Artificial Intelligence. Philadelphia, v. 31, n. 9-10, p. 764-791, 2017.
ISSN 0883-9514 (Sherpa/Romeo, impact factor)
Publisher Taylor & Francis Inc
Extent 764-791
Origin http://dx.doi.org/10.1080/08839514.2018.1444334
Access rights Closed access
Type Article
Web of Science ID WOS:000432145400005
URI https://repositorio.unifesp.br/handle/11600/55279

Show full item record


File Size Format View

There are no files associated with this item.

This item appears in the following Collection(s)




My Account