Fuzzy joint mutual information feature selection based on ideal vector

作者：

Highlights：

• A new Feature Selection, called FJMIIV, is proposed for classification models.

• FJMIIV uses fuzzy mutual information measures based on ideal vector.

• FJMIIV depends on feature-idealvector instead of feature-feature to save the cost.

• FJMIIV considers the joint discriminative ability to improve the feature selection.

• FJMIIV outperforms some state-of-the-art methods in the classification performance.

摘要

•A new Feature Selection, called FJMIIV, is proposed for classification models.•FJMIIV uses fuzzy mutual information measures based on ideal vector.•FJMIIV depends on feature-idealvector instead of feature-feature to save the cost.•FJMIIV considers the joint discriminative ability to improve the feature selection.•FJMIIV outperforms some state-of-the-art methods in the classification performance.

论文关键词：Feature selection,Fuzzy sets,Information theory,Ideal vector,Classification models

论文评审过程：Received 20 April 2021, Revised 18 December 2021, Accepted 23 December 2021, Available online 14 January 2022, Version of Record 20 January 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116453