首页 | 本学科首页   官方微博 | 高级检索  
     

迭代二叉树3代算法的一种改进
引用本文:朱金坛.迭代二叉树3代算法的一种改进[J].西安邮电学院学报,2014(1):37-41.
作者姓名:朱金坛
作者单位:西安铁路职业技术学院电子信息系,陕西西安710014
基金项目:陕西省自然科学基金资助项目(2012JQ8050)
摘    要:针对数据挖掘决策树中迭代二叉树3代算法复杂的对数运算以及属性取值多向依赖的缺陷,提出了一种改进算法。该算法将对数运算改进为简易的普通运算,引入重要度、关联度概念以及调整系数,形成一个综合评价指数来确定作为决策树生成的划分结点的属性。仿真结果表明,改进算法简化了计算过程、提高了运算效率,同时使得决策树的形成不依赖属性多值取向。

关 键 词:迭代二叉树3代算法(ID3)  决策树  粗糙集论  重要度  关联度

An improved algorithm of iterative dichotomiser 3
ZHU Jintan.An improved algorithm of iterative dichotomiser 3[J].Journal of Xi'an Institute of Posts and Telecommunications,2014(1):37-41.
Authors:ZHU Jintan
Affiliation:ZHU Jintan (Department of Electronic Information, XVan Railway Vocational and Technical Institute, XVan 710014, China)
Abstract:Iterative dichotomiser 3 (ID3) algorithm for data mining of complex logarithmic and property values more dependence on defect, This paper presents an improved algorithm. The difficult part of the logarithmic in the original algorithm is improved into an easy one. At the same time, the importance, the concept of correlation degree and the adjustment of the coefficient are introduced in this article. Finally, a comprehensive evaluation index is formed to determine the choice of a property as the division of the decision tree generated notes. Simulation result show that the improved algorithm can simplify the calculation, improve the efficiency of the operation, and make the formation of the decision tress independent of the formation of attribute value orientation.
Keywords:Iterative Dichotomiser 3 (ID3)  decision tree  rough sets  importance degree  correlation degree
本文献已被 CNKI 维普 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号