您的位置:山东大学 -> 科技期刊社 -> 《山东大学学报(理学版)》

J4 ›› 2010, Vol. 45 ›› Issue (7): 114-118.

• 论文 • 上一篇    下一篇

决策树构建方法:向前两步优于一步

张雯,张化祥*,李明方,计华   

  1. 山东师范大学信息科学与工程学院,  山东 济南 250014
  • 收稿日期:2010-04-02 出版日期:2010-07-16 发布日期:2010-09-06
  • 通讯作者: 张化祥(1966-),男,教授,博士生导师,主要研究方向为机器学习、模式识别及Web挖掘等.
  • 作者简介:张雯(1984-),女,硕士研究生,主要研究方向为数据挖掘、机器学习. Email:zhangwen0927@126.com
  • 基金资助:

    山东省科技研究计划项目(2007ZZ17,2008GG10001015,2008B0026);山东省教育厅科研项目(J09LG02)

A decision tree construction approach: two-step forward is better than one

ZHANG Wen, ZHANG Hua-xiang*, LI Ming-fang, JI Hua   

  1. School of Information Science and Engineering, Shandong Normal University, Jinan 250014, Shandong, China
  • Received:2010-04-02 Online:2010-07-16 Published:2010-09-06

摘要:

为提高搜索算法找到全局最优解的可能性,在C4.5算法的基础上,本文提出了向前两步的决策树(two-step forward decision tree,TSFDT)构建算法。该算法在选择属性时,考虑同时选择两个属性带来的信息增益,而不是只考虑单一最优属性对于信息增益的贡献,从而在寻找问题全局最优方面比只考虑单一最优属性具有更大的可能性。10个UCI基准数据集上的实验结果表明,该算法明显优于C4.5算法。

关键词: 决策树;信息增益;C4.5算法;局部最优

Abstract:

In order to increase the probability of finding the global optimum, a novel decision tree construction algorithm adopting two-step forward idea was proposed based on C4.5 algorithm. The algorithm  was more possible to get the global optimum of a classification task because it  considered  the information gained from  selecting two attributes simultaneously, rather than the information gained from  just selecting an optimal single attribute. Experimental results on 10 UCI benchmark data sets showed that it outperforms C4.5.

Key words: decision tree; information gain; C4.5 algorithm; local optimum

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!