A novel approach to pre-extracting support vectors based on the theory of belief functions

作者：

Highlights：

• A new approach to pre-extracting support vectors for SVM classifiers is proposed.

• The new approach is based on the theory of belief functions.

• The new approach can detect the noisy samples and outliers.

• The new approach can reduce SVM learning cost without classification accuracy loss.

摘要

•A new approach to pre-extracting support vectors for SVM classifiers is proposed.•The new approach is based on the theory of belief functions.•The new approach can detect the noisy samples and outliers.•The new approach can reduce SVM learning cost without classification accuracy loss.

论文关键词：SVM,Belief functions,Pre-extraction,Pattern recognition

论文评审过程：Received 3 February 2016, Revised 11 June 2016, Accepted 21 July 2016, Available online 22 July 2016, Version of Record 29 September 2016.

论文官网地址：https://doi.org/10.1016/j.knosys.2016.07.029