A PSO-based multi-objective multi-label feature selection method in classification

Abstract Feature selection is an important data preprocessing technique in multi-label classification. Although a large number of studies have been proposed to tackle feature selection problem, there are a few cases for multi-label data. This paper studies a multi-label feature selection algorithm u...

Full description

Bibliographic Details
Main Authors:	Yong Zhang, Dun-wei Gong, Xiao-yan Sun, Yi-nan Guo
Format:	Article
Language:	English
Published:	Nature Publishing Group 2017-03-01
Series:	Scientific Reports
Online Access:	https://doi.org/10.1038/s41598-017-00416-0

id	doaj-0d11cab47de64ebd87afaa63a71b6192
record_format	Article
spelling	doaj-0d11cab47de64ebd87afaa63a71b61922020-12-08T02:25:43ZengNature Publishing GroupScientific Reports2045-23222017-03-017111210.1038/s41598-017-00416-0A PSO-based multi-objective multi-label feature selection method in classificationYong Zhang0Dun-wei Gong1Xiao-yan Sun2Yi-nan Guo3School of Information and Electronic Engineering, China University of Mining and TechnologySchool of Information and Electronic Engineering, China University of Mining and TechnologySchool of Information and Electronic Engineering, China University of Mining and TechnologySchool of Information and Electronic Engineering, China University of Mining and TechnologyAbstract Feature selection is an important data preprocessing technique in multi-label classification. Although a large number of studies have been proposed to tackle feature selection problem, there are a few cases for multi-label data. This paper studies a multi-label feature selection algorithm using an improved multi-objective particle swarm optimization (PSO), with the purpose of searching for a Pareto set of non-dominated solutions (feature subsets). Two new operators are employed to improve the performance of the proposed PSO-based algorithm. One operator is adaptive uniform mutation with action range varying over time, which is used to extend the exploration capability of the swarm; another is a local learning strategy, which is designed to exploit the areas with sparse solutions in the search space. Moreover, the idea of the archive, and the crowding distance are applied to PSO for finding the Pareto set. Finally, experiments verify that the proposed algorithm is a useful approach of feature selection for multi-label classification problem.https://doi.org/10.1038/s41598-017-00416-0
collection	DOAJ
language	English
format	Article
sources	DOAJ
author	Yong Zhang Dun-wei Gong Xiao-yan Sun Yi-nan Guo
spellingShingle	Yong Zhang Dun-wei Gong Xiao-yan Sun Yi-nan Guo A PSO-based multi-objective multi-label feature selection method in classification Scientific Reports
author_facet	Yong Zhang Dun-wei Gong Xiao-yan Sun Yi-nan Guo
author_sort	Yong Zhang
title	A PSO-based multi-objective multi-label feature selection method in classification
title_short	A PSO-based multi-objective multi-label feature selection method in classification
title_full	A PSO-based multi-objective multi-label feature selection method in classification
title_fullStr	A PSO-based multi-objective multi-label feature selection method in classification
title_full_unstemmed	A PSO-based multi-objective multi-label feature selection method in classification
title_sort	pso-based multi-objective multi-label feature selection method in classification
publisher	Nature Publishing Group
series	Scientific Reports
issn	2045-2322
publishDate	2017-03-01
description	Abstract Feature selection is an important data preprocessing technique in multi-label classification. Although a large number of studies have been proposed to tackle feature selection problem, there are a few cases for multi-label data. This paper studies a multi-label feature selection algorithm using an improved multi-objective particle swarm optimization (PSO), with the purpose of searching for a Pareto set of non-dominated solutions (feature subsets). Two new operators are employed to improve the performance of the proposed PSO-based algorithm. One operator is adaptive uniform mutation with action range varying over time, which is used to extend the exploration capability of the swarm; another is a local learning strategy, which is designed to exploit the areas with sparse solutions in the search space. Moreover, the idea of the archive, and the crowding distance are applied to PSO for finding the Pareto set. Finally, experiments verify that the proposed algorithm is a useful approach of feature selection for multi-label classification problem.
url	https://doi.org/10.1038/s41598-017-00416-0
work_keys_str_mv	AT yongzhang apsobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT dunweigong apsobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT xiaoyansun apsobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT yinanguo apsobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT yongzhang psobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT dunweigong psobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT xiaoyansun psobasedmultiobjectivemultilabelfeatureselectionmethodinclassification AT yinanguo psobasedmultiobjectivemultilabelfeatureselectionmethodinclassification
_version_	1724393681769725952

A PSO-based multi-objective multi-label feature selection method in classification

Similar Items