Employing Linear Regression in Regression Tree Leaves
Employing Linear Regression in Regression Tree Leaves
复制标题
在回归树叶中采用线性回归
DOI：
--
复制
发表时间：
1992
期刊：
影响因子：
--
通讯作者：
Aram Karalic
中科院分区：
文献类型：
--
作者：
Aram Karalic
本文分析了在回归树的叶节点上使用线性回归的优点。讨论了这种修改如何影响回归树的构造、剪枝和解释。该改进在人工和现实领域进行了测试。结果表明，修正后的回归树分类误差较小，是有益的。
The advantage of using linear regression in the leaves of a regression tree is analysed in the paper. It is carried out how this modiication aaects the construction, pruning and interpretation of a regression tree. The modiication is tested on artiicial and real-life domains. The results show that the mod-iication is beneecial as it leads to smaller classiica-tion errors of induced regression trees.