Broaden the minority class space for decision tree induction using antigen-derived detectors

作者：

Highlights：

• This paper focuses on broadening the minority space by learning both majority class space and minority class space.

• A negative selection over-sampling technology (NSOTE) is proposed.

• Previous over-sampling methods only learn minority class space to produce minority class examples.

• We also investigate the performance of NSOTE and previous over-sampling methods on artificial and real datasets.

摘要

•This paper focuses on broadening the minority space by learning both majority class space and minority class space.•A negative selection over-sampling technology (NSOTE) is proposed.•Previous over-sampling methods only learn minority class space to produce minority class examples.•We also investigate the performance of NSOTE and previous over-sampling methods on artificial and real datasets.

论文关键词：Negative selection,Imbalance learning,Resampling,Decision tree

论文评审过程：Received 28 February 2017, Revised 17 September 2017, Accepted 21 September 2017, Available online 28 September 2017, Version of Record 18 October 2017.

论文官网地址：https://doi.org/10.1016/j.knosys.2017.09.029