A two-dimensional (2-D) learning framework for Particle Swarm based feature selection

作者：

Highlights：

• A new learning method is proposed for PSO based feature selection.

• Subset cardinality is used as the extra learning dimension to improve the search.

• Position update process through inclusive learning on cardinality and feature.

• Key elements of canonical PSO are retained despite the extra learning dimension.

摘要

•A new learning method is proposed for PSO based feature selection.•Subset cardinality is used as the extra learning dimension to improve the search.•Position update process through inclusive learning on cardinality and feature.•Key elements of canonical PSO are retained despite the extra learning dimension.

论文关键词：Classification,Dimensionality reduction,Feature selection,Particle Swarm Optimization,Machine learning

论文评审过程：Received 4 November 2016, Revised 1 October 2017, Accepted 21 November 2017, Available online 22 November 2017, Version of Record 1 December 2017.

论文官网地址：https://doi.org/10.1016/j.patcog.2017.11.027