Fuzzy filter cost-sensitive feature selection with differential evolution

作者：

Highlights：

• A cost sensitive filter evaluation criterion is developed.

• Based on the filter evaluation criterion, a EC-based cost-sensitive feature selection method is developed.

• It is expected to effectively search for lower-cost features, which are beneficial for the learning performance.

• The proposed cost-sensitive method also reduces the computational cost and the total misclassification cost.

摘要

•A cost sensitive filter evaluation criterion is developed.•Based on the filter evaluation criterion, a EC-based cost-sensitive feature selection method is developed.•It is expected to effectively search for lower-cost features, which are beneficial for the learning performance.•The proposed cost-sensitive method also reduces the computational cost and the total misclassification cost.

论文关键词：Feature selection,Cost-sensitive,Classification,Differential evolution

论文评审过程：Received 9 May 2020, Revised 19 January 2022, Accepted 19 January 2022, Available online 25 January 2022, Version of Record 8 February 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.108259