Stable feature selection using copula based mutual information
作者：
Highlights：
• A novel method for feature selection using incorporation of copula based multivariate dependency in mutual information, which assists to remove the need to average out over multiple instances of bivariate dependencies.
• The method is unbiased against noisy datasets due to the scale-invariant property of copula.
• The method can be applied to datasets, in which the ratio between sample size to class size is large enough, even though original marginal distributions are unknown.
• The method also satisfies the maximum relevance and minimum redundancy criteria of feature selection.
摘要
•A novel method for feature selection using incorporation of copula based multivariate dependency in mutual information, which assists to remove the need to average out over multiple instances of bivariate dependencies.•The method is unbiased against noisy datasets due to the scale-invariant property of copula.•The method can be applied to datasets, in which the ratio between sample size to class size is large enough, even though original marginal distributions are unknown.•The method also satisfies the maximum relevance and minimum redundancy criteria of feature selection.
论文关键词：Copula,Feature selection,Mutual information,Stability,Classification accuracy
论文评审过程：Received 5 February 2020, Revised 3 October 2020, Accepted 7 October 2020, Available online 26 October 2020, Version of Record 30 January 2021.
论文官网地址：https://doi.org/10.1016/j.patcog.2020.107697
