Two approaches for novelty detection using random forest

作者：

Highlights：

• A framework for novelty detection using random forest is proposed.

• Two specific approaches using the vote distribution and the proximity matrix are presented.

• A comprehensive empirical study on both synthetic and real-world datasets is conducted.

摘要

•A framework for novelty detection using random forest is proposed.•Two specific approaches using the vote distribution and the proximity matrix are presented.•A comprehensive empirical study on both synthetic and real-world datasets is conducted.

论文关键词：Novelty detection,Ensemble learning,Random forest,Proximity matrix

论文评审过程：Available online 19 February 2015.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.12.028