Semi-supervised regression trees with application to QSAR modelling

作者：

Highlights：

• Obtaining labelled data for many domains is a very difficult and expensive task.

• Semi-supervised learning leverages the information from labelled and unlabelled data.

• The proposed semi-supervised regression trees outperform supervised regression trees.

• Semi-supervised regression trees can be easily applied to QSAR modelling.

• Semi-supervised regression trees are readily interpretable models.

摘要

•Obtaining labelled data for many domains is a very difficult and expensive task.•Semi-supervised learning leverages the information from labelled and unlabelled data.•The proposed semi-supervised regression trees outperform supervised regression trees.•Semi-supervised regression trees can be easily applied to QSAR modelling.•Semi-supervised regression trees are readily interpretable models.

论文关键词：Semi-supervised learning,Regression,Decision trees,Random forests,QSAR

论文评审过程：Received 11 July 2019, Revised 18 February 2020, Accepted 13 May 2020, Available online 21 May 2020, Version of Record 9 June 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.113569