| 决策树构建方法:向前两步优于一步 |
| |
| 引用本文： | 张雯,张化祥,李明方,计华. 决策树构建方法:向前两步优于一步[J]. 山东大学学报(理学版), 2010, 45(7): 114-118 |
| |
| 作者姓名： | 张雯 张化祥 李明方 计华 |
| |
| 作者单位： | 山东师范大学信息科学与工程学院, 山东 济南 250014 |
| |
| 基金项目： | 山东省科技研究计划项目，山东省教育厅科研项目 |
| |
| 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, |
| 收稿时间： | 2010-04-02 |
A decision tree construction approach: two-step forward is better than one |
| |
| ZHANG Wen,ZHANG Hua-xiang,LI Ming-fang,JI Hua. A decision tree construction approach: two-step forward is better than one[J]. Journal of Shandong University, 2010, 45(7): 114-118 |
| |
| Authors: | ZHANG Wen ZHANG Hua-xiang LI Ming-fang JI Hua |
| |
| Affiliation: | School of Information Science and Engineering, Shandong Normal University, Jinan 250014, Shandong, China |
| |
| 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. |
| |
| Keywords: | decision tree information gain C4.5 algorithm local optimum |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《山东大学学报(理学版)》浏览原始摘要信息 |
|
点击此处可从《山东大学学报(理学版)》下载全文 |
|
