Simultaneous feature selection and discretization based on mutual information
作者：
Highlights：
• We address discretization and feature selection jointly with a single criteria.
• The proposed discretization method is dynamic and independent of classification algorithms.
• The amount of errors introduced for Relevancy, Redundancy and Complementary Information are derived analytically.
• It is also analytically shown that Relevancy, Redundancy and Complementary follows χ2-distribution.
• A χ2-based search is introduced to select a small set of features and to discretize them with small number of intervals.
摘要
•We address discretization and feature selection jointly with a single criteria.•The proposed discretization method is dynamic and independent of classification algorithms.•The amount of errors introduced for Relevancy, Redundancy and Complementary Information are derived analytically.•It is also analytically shown that Relevancy, Redundancy and Complementary follows χ2-distribution.•A χ2-based search is introduced to select a small set of features and to discretize them with small number of intervals.
论文关键词：Feature selection,Mutual information,Bias,Dynamic discretization
论文评审过程：Received 16 August 2017, Revised 14 February 2019, Accepted 19 February 2019, Available online 19 February 2019, Version of Record 27 February 2019.
论文官网地址：https://doi.org/10.1016/j.patcog.2019.02.016
