OFS-Density: A novel online streaming feature selection method

作者：

Highlights：

• We propose a novel online feature selection method (OFS-Density) which does not require domain information before learning.

• We proposed a new neighborhood relation which makes OFS-Density need not specify any parameters in advance.

• OFS-Density uses a fuzzy equal constraint for redundant analysis.

• The experiment results demonstrate the efficiency of our new methods.

摘要

•We propose a novel online feature selection method (OFS-Density) which does not require domain information before learning.•We proposed a new neighborhood relation which makes OFS-Density need not specify any parameters in advance.•OFS-Density uses a fuzzy equal constraint for redundant analysis.•The experiment results demonstrate the efficiency of our new methods.

论文关键词：Feature selection,Online feature selection,Streaming features,Neighborhood rough set

论文评审过程：Received 13 December 2017, Revised 20 June 2018, Accepted 27 August 2018, Available online 1 September 2018, Version of Record 10 September 2018.

论文官网地址：https://doi.org/10.1016/j.patcog.2018.08.009