Optimizing multi-objective PSO based feature selection method using a feature elitism mechanism
作者：
Highlights：
• A multi-objective PSO-based feature selection method called RSPSOFS is proposed.
• RSPSOFS proposes a feature-elitism mechanism based on the frequency of features.
• Problem space information improves the evolutionary process of RSPSOFS.
• Archive refinement and purposeful particles movement are the important achievements.
• Qualitative and quantitative analyses of the results confirm the performance of RFPSOFS.
摘要
•A multi-objective PSO-based feature selection method called RSPSOFS is proposed.•RSPSOFS proposes a feature-elitism mechanism based on the frequency of features.•Problem space information improves the evolutionary process of RSPSOFS.•Archive refinement and purposeful particles movement are the important achievements.•Qualitative and quantitative analyses of the results confirm the performance of RFPSOFS.
论文关键词：Multi-objective,Particle swarm optimization,Feature selection,Feature ranking,Feature elitism
论文评审过程：Received 23 January 2018, Revised 13 April 2018, Accepted 3 July 2018, Available online 11 July 2018, Version of Record 20 July 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.07.013
