SI(FS)2: Fast simultaneous instance and feature selection for datasets with many features

作者：

Highlights：

• We propose a new simultaneous instance and feature selection algorithm.

• The method achieves better storage reduction and testing error than previous approaches.

• The method is scalable to datasets with millions of features.

摘要

•We propose a new simultaneous instance and feature selection algorithm.•The method achieves better storage reduction and testing error than previous approaches.•The method is scalable to datasets with millions of features.

论文关键词：Instance selection,Feature selection,Evolutionary algorithms,Knearest neighbor rule

论文评审过程：Received 6 May 2020, Revised 10 August 2020, Accepted 23 October 2020, Available online 24 October 2020, Version of Record 2 November 2020.

论文官网地址：https://doi.org/10.1016/j.patcog.2020.107723