Enhanced NSGA-II-based feature selection method for high-dimensional classification

被引:25
作者
Li, Min [1 ,2 ]
Ma, Huan [1 ]
Lv, Siyu [1 ]
Wang, Lei [1 ]
Deng, Shaobo [1 ]
机构
[1] Nanchang Inst Technol, Sch Informat Engn, Nanchang 330099, Peoples R China
[2] Nanchang Inst Technol, 289 Tianxiang Ave, Nanchang 330099, Peoples R China
基金
中国国家自然科学基金;
关键词
NSGA-II; Feature selection; Multi -objective optimization; High -dimensional data; Classification; MULTIOBJECTIVE FEATURE-SELECTION; PARTICLE SWARM OPTIMIZATION; GENETIC ALGORITHM;
D O I
10.1016/j.ins.2024.120269
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Feature selection in high-dimensional data faces significant challenges owing to large and discrete decision spaces. In this study, we propose a feature selection method based on the nondominated sorting genetic algorithm-II (NSGA-II) to enhance the performance of feature selection in highdimensional data. This study makes four contributions: 1) The sparse initialization strategy is used to sparsen the search space and accelerate the convergence speed of the algorithm; 2) the guided selection operator is employed to strike a balance between exploration and exploitation abilities; 3) an intra-population evolution-based mutation operator dynamically shrinks the search space; and 4) a greedy repair strategy is adopted to generate improved feature subsets. The proposed method was validated on 15 publicly available high-dimensional datasets and compared with eight competitive multi-objective feature selection methods. The results demonstrate that the proposed method can achieve superior classification accuracy in a shorter time, with a smaller subset of features containing less redundancy.
引用
收藏
页数:29
相关论文
共 50 条
[41]   Differential Evolution-Based Feature Selection: A Niching-Based Multiobjective Approach [J].
Wang, Peng ;
Xue, Bing ;
Liang, Jing ;
Zhang, Mengjie .
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2023, 27 (02) :296-310
[42]   A Grid-dominance based Multi-objective Algorithm for Feature Selection in Classification [J].
Wang, Peng ;
Xue, Bing ;
Zhang, Mengjie ;
Liang, Jing .
2021 IEEE CONGRESS ON EVOLUTIONARY COMPUTATION (CEC 2021), 2021, :2053-2060
[43]   Feature Selection with Integrated Relevance and Redundancy Optimization [J].
Xu, Linli ;
Zhou, Qi ;
Huang, Aiqing ;
Ouyang, Wenjun ;
Chen, Enhong .
2015 IEEE INTERNATIONAL CONFERENCE ON DATA MINING (ICDM), 2015, :1063-1068
[44]   A Survey on Evolutionary Computation Approaches to Feature Selection [J].
Xue, Bing ;
Zhang, Mengjie ;
Browne, Will N. ;
Yao, Xin .
IEEE TRANSACTIONS ON EVOLUTIONARY COMPUTATION, 2016, 20 (04) :606-626
[45]   Particle Swarm Optimization for Feature Selection in Classification: A Multi-Objective Approach [J].
Xue, Bing ;
Zhang, Mengjie ;
Browne, Will N. .
IEEE TRANSACTIONS ON CYBERNETICS, 2013, 43 (06) :1656-1671
[46]   A bidirectional dynamic grouping multi-objective evolutionary algorithm for feature selection on high-dimensional classification [J].
Yu, Kunjie ;
Sun, Shaoru ;
Liang, Jing ;
Chen, Ke ;
Qu, Boyang ;
Yue, Caitong ;
Wang, Ling .
INFORMATION SCIENCES, 2023, 648
[47]   Adaptive Semi-Supervised Classifier Ensemble for High Dimensional Data Classification [J].
Yu, Zhiwen ;
Zhang, Yidong ;
You, Jane ;
Chen, C. L. Philip ;
Wong, Hau-San ;
Han, Guoqiang ;
Zhang, Jun .
IEEE TRANSACTIONS ON CYBERNETICS, 2019, 49 (02) :366-379
[48]  
Yue CT, 2019, IEEE C EVOL COMPUTAT, P302, DOI [10.1109/cec.2019.8790329, 10.1109/CEC.2019.8790329]
[49]   Multi-objective optimization algorithm based on clustering guided binary equilibrium optimizer and NSGA-III to solve high-dimensional feature selection problem [J].
Zhang, Min ;
Wang, Jie-Sheng ;
Liu, Yu ;
Song, Hao-Ming ;
Hou, Jia-Ning ;
Wang, Yu-Cai ;
Wang, Min .
INFORMATION SCIENCES, 2023, 648
[50]   Binary differential evolution with self-learning for multi-objective feature selection [J].
Zhang, Yong ;
Gong, Dun-wei ;
Gao, Xiao-zhi ;
Tian, Tian ;
Sun, Xiao-yan .
INFORMATION SCIENCES, 2020, 507 :67-85