Random forest for label ranking

作者：

Highlights：

• An effective random forest based label ranking method is proposed.

• A novel two-step rank aggregation strategy is proposed.

• The proposed method is evaluated on benchmarks with complete and partial ranking.

• The proposed method is highly competitive compared with state-of-the-art methods.

摘要

•An effective random forest based label ranking method is proposed.•A novel two-step rank aggregation strategy is proposed.•The proposed method is evaluated on benchmarks with complete and partial ranking.•The proposed method is highly competitive compared with state-of-the-art methods.

论文关键词：Preference learning,Label ranking,Random forest,Decision tree

论文评审过程：Received 30 October 2017, Revised 14 June 2018, Accepted 15 June 2018, Available online 18 June 2018, Version of Record 26 June 2018.

论文官网地址：https://doi.org/10.1016/j.eswa.2018.06.036