Feature selection using bare-bones particle swarm optimization with mutual information

作者：

Highlights：

• Proposing a novel PSO-based feature selection algorithm with mutual information.

• Presenting an effective swarm initialization strategy based on label correlation.

• Designing two local search operators, the supplementary and deletion operators.

• Giving an adaptive flip mutation to help particles jump out of local extremum.

摘要

•Proposing a novel PSO-based feature selection algorithm with mutual information.•Presenting an effective swarm initialization strategy based on label correlation.•Designing two local search operators, the supplementary and deletion operators.•Giving an adaptive flip mutation to help particles jump out of local extremum.

论文关键词：Feature selection,Particle swarm,Swarm initialization,Mutual information,Local search

论文评审过程：Received 4 December 2019, Revised 14 September 2020, Accepted 22 December 2020, Available online 31 December 2020, Version of Record 7 January 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107804