Exploring interactive attribute reduction via fuzzy complementary entropy for unlabeled mixed data

作者：

Highlights：

• Fuzzy complementary entropy is employed to explore unsupervised interactive attribute reduction.

• A novel evaluation index of attribute importance is constructed for selecting mixed attribute by using the idea of unsupervised maximum information-minimum redundancy-maximum interactivity.

• The experimental results show that the proposed algorithm has better performance.

摘要

•Fuzzy complementary entropy is employed to explore unsupervised interactive attribute reduction.•A novel evaluation index of attribute importance is constructed for selecting mixed attribute by using the idea of unsupervised maximum information-minimum redundancy-maximum interactivity.•The experimental results show that the proposed algorithm has better performance.

论文关键词：Fuzzy rough set theory,Unsupervised attribute reduction,Complementary entropy,Maximal information,Minimal redundancy,Maximal interactivity,Mixed data

论文评审过程：Received 12 February 2021, Revised 7 March 2022, Accepted 12 March 2022, Available online 15 March 2022, Version of Record 21 March 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.108651