A Feature Selection based on perturbation theory

作者：

Highlights：

• We have proved that perturbation theory can detect correlations between features.

• PFS chooses a fraction of the number of features selected by other methods.

• This is the first work to report on using perturbation theory in feature selection.

• The complexity of PFS is dominated by the complexity of SVD O(min{mn2, m2n}).

摘要

•We have proved that perturbation theory can detect correlations between features.•PFS chooses a fraction of the number of features selected by other methods.•This is the first work to report on using perturbation theory in feature selection.•The complexity of PFS is dominated by the complexity of SVD O(min{mn2, m2n}).

论文关键词：Feature selection,Perturbation theory,Least angle regression

论文评审过程：Received 29 September 2018, Revised 30 January 2019, Accepted 20 February 2019, Available online 5 March 2019, Version of Record 5 March 2019.

论文官网地址：https://doi.org/10.1016/j.eswa.2019.02.028