Linear classifier design under heteroscedasticity in Linear Discriminant Analysis

作者：

Highlights：

• We derive a linear classifier for heteroscedastic linear discriminant analysis.

• The proposed scheme efficiently minimises the Bayes error for binary classification.

• A local neighbourhood search is also proposed for non-normal distributions.

• The proposed schemes are experimentally validated on twelve datasets.

摘要

•We derive a linear classifier for heteroscedastic linear discriminant analysis.•The proposed scheme efficiently minimises the Bayes error for binary classification.•A local neighbourhood search is also proposed for non-normal distributions.•The proposed schemes are experimentally validated on twelve datasets.

论文关键词：LDA,Heteroscedasticity,Bayes error,Linear classifier

论文评审过程：Received 12 November 2016, Revised 23 February 2017, Accepted 24 February 2017, Available online 24 February 2017, Version of Record 27 March 2017.

论文官网地址：https://doi.org/10.1016/j.eswa.2017.02.039