Collaboration graph for feature set partitioning in data classification

作者：

Highlights：

• A measure defined to show the effectiveness of each two features in classification.

• Collaboration Graph (CG) represents the measure as an edge between each two features.

• Community detection is used on CG to specify informative feature subsets.

• The approach has been tested successfully on real and synthetic data.

摘要

•A measure defined to show the effectiveness of each two features in classification.•Collaboration Graph (CG) represents the measure as an edge between each two features.•Community detection is used on CG to specify informative feature subsets.•The approach has been tested successfully on real and synthetic data.

论文关键词：Ensemble Classification,Features Collaboration Graph,Community Detection,AdaBoost Algorithm

论文评审过程：Received 21 June 2022, Revised 7 September 2022, Accepted 4 October 2022, Available online 12 October 2022, Version of Record 19 October 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118988