节点文献
基于粗集理论的决策树剪枝
RST in Decision Tree Pruning
【摘要】 本文在理论上分析比较了基于粗糙集理论的剪枝方法和C4.5中的EBP剪枝方法,并通过在多个数据集上进行实验比较,证实了基于粗糙集理论剪枝方法的优越性。
【Abstract】 In this paper, the decision tree pruning method based on RST is compared with the Error Based Pruning method used in C4.5 in theory, and they are compared by tests on some databases. The result proves the validity of the pruning method based on RST.
【关键词】 过匹配;
剪枝;
深度拟合率;
错误率;
决策树;
【Key words】 over-fitting; pruning; depth-fitting ratio; error ratio; decision tree;
【Key words】 over-fitting; pruning; depth-fitting ratio; error ratio; decision tree;
【基金】 吉林省自然科学基金资助项目(20040529);东北师范大学校内青年基金资助项目(111420000)
- 【文献出处】 计算机工程与科学 ,Computer Engineering & Science , 编辑部邮箱 ,2007年01期
- 【分类号】TP18
- 【被引频次】10
- 【下载频次】263