首页 | 本学科首页   官方微博 | 高级检索  
     

一种健壮有效的决策树改进模型
引用本文:卜亚杰,胡朝举,白兰,董娜.一种健壮有效的决策树改进模型[J].计算机应用,2008,28(Z1).
作者姓名:卜亚杰  胡朝举  白兰  董娜
作者单位:1. 华北电力大学计算机科学与技术学院,河北,保定,071003
2. 华北电力大学计算机系教研室,河北,保定,071003
摘    要:提出了一种健壮有效的决策树改进模型AJD3.该决策树模型基于经典的ID3决策树模型,在属性的选取上进行了改进.利用引入属性优先关联度参数计算节点的修正信息增益,并选择具有最高修正增益的属性作为当前节点的测试属性.实验表明,AID3决策树模型在提高分类准确率的同时,有效地增强了模型的健壮性.

关 键 词:决策树  ID3  属性优先关联度  AID3

Improved model for robust and effective decision tree
BU Ya-jie,HU Chao-ju,BAI Lan,DONG Na.Improved model for robust and effective decision tree[J].journal of Computer Applications,2008,28(Z1).
Authors:BU Ya-jie  HU Chao-ju  BAI Lan  DONG Na
Affiliation:BU Ya-jie1,HU Chao-ju2,BAI Lan1,DONG Na1(1.School of Computer Science , Technology,North China Electric Power University,Baoding Hebei 071003,China,2.Laboratory of Computer,China)
Abstract:The robust and effective decision tree improved model AID3 was introduced.The model was based on ID3,but it was improved in attribute selection.The modified information gain of certain node was calculated by using attribution value association parameter,and attributed with the highest information gain was chosen as the test attribute for the current node.The experiment has proved that AID3 decision tree model improved the accuracy and robustness of classification.
Keywords:decision tree  ID3  attribute value association  AID3  
本文献已被 CNKI 万方数据 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