A weighted information-gain measure for ordinal classification trees
作者：
Highlights：
• We define an ordinal oriented information metric, based on weighted entropy.
• We propose an ordinal oriented decision-tree, using the new information metric.
• The new decision-tree method is effective for ordinal classification problems.
• The new tree outperforms C4.5 on most datasets with an ordinal target.
• The new tree outperforms Random Forest on several datasets with an ordinal target.
摘要
•We define an ordinal oriented information metric, based on weighted entropy.•We propose an ordinal oriented decision-tree, using the new information metric.•The new decision-tree method is effective for ordinal classification problems.•The new tree outperforms C4.5 on most datasets with an ordinal target.•The new tree outperforms Random Forest on several datasets with an ordinal target.
论文关键词：Information-gain,Decision trees,Classification tree,Weighted entropy,C4.5,Ordinal classification
论文评审过程：Received 7 August 2019, Revised 23 February 2020, Accepted 10 March 2020, Available online 13 March 2020, Version of Record 20 March 2020.
论文官网地址：https://doi.org/10.1016/j.eswa.2020.113375
