Statistical-mean double-quantitative K-nearest neighbor classification learning based on neighborhood distance measurement

作者：

Highlights：

• Construct arithmetic-mean, geometric-mean double-quantitative distances, classifiers.

• 2 distances offer statistic measurement, information fusion, good property, function.

• 2 new classifiers improve current KNN, KNGR, KNGA, KNGD to gain optimal performances.

• Promote neighborhood measurement and classification learning by double quantization.

摘要

•Construct arithmetic-mean, geometric-mean double-quantitative distances, classifiers.•2 distances offer statistic measurement, information fusion, good property, function.•2 new classifiers improve current KNN, KNGR, KNGA, KNGD to gain optimal performances.•Promote neighborhood measurement and classification learning by double quantization.

论文关键词：Neighborhood rough sets,Granular computing,Statistical distance measurement,Double quantization,Classification learning,K-nearest neighbor classifier

论文评审过程：Received 30 November 2021, Revised 27 March 2022, Accepted 9 May 2022, Available online 19 May 2022, Version of Record 26 May 2022.

论文官网地址：https://doi.org/10.1016/j.knosys.2022.109018