| 基于粗集的混合变量决策树构造算法研究 |
| |
| 引用本文： | 胡学钢,张冬艳.基于粗集的混合变量决策树构造算法研究[J].合肥工业大学学报(自然科学版),2007,30(3):257-260. |
| |
| 作者姓名： | 胡学钢 张冬艳 |
| |
| 作者单位： | 合肥工业大学,计算机与信息学院,安徽,合肥,230009 |
| |
| 摘 要： | 文章提出混合变量决策树结构,并在此基础上提出基于粗集理论的混合变量决策树构造算法RSH2,算法在每个结点选择尽可能少的属性明确划分尽可能多的实例,减小了决策树规模,且易于理解。将RSH2算法与ID3算法及基于粗集的单变量决策树算法HACRs进行实验比较,结果表明该算法有良好性能。
|
| 关 键 词： | 单变量决策树 多变量决策树 粗糙集合 归纳学习 |
| 文章编号： | 1003-5060(2007)03-0257-04 |
| 修稿时间： | 2006年2月21日 |
On the decision tree inductive algorithm based on the rough set theory |
| |
| HU Xue-gang,ZHANG Dong-yan.On the decision tree inductive algorithm based on the rough set theory[J].Journal of Hefei University of Technology(Natural Science),2007,30(3):257-260. |
| |
| Authors: | HU Xue-gang ZHANG Dong-yan |
| |
| Abstract: | The structure of the hybrid decision tree and the constructing algorithm for the structure are proposed in this paper.The hybrid decision tree algorithm RSH2,which is based on the rough set theory,selects the attributes as few as possible which can classify the instants as many as possible.Thus,the scale of the decision tree will be diminished and the tree will be easier to understand.In ad-dition,the conditional information entropy based reduction algorithm is used before constructing the tree so that time will be saved.The comparison among the algorithms RSH2,ID3 and HACRs based on the rough set is made with experiments,and the RSH2 algorithm is proved to have good perform-ance. |
| |
| Keywords: | univariate decision tree multivariate decision tree rough set inductive learning |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
