An Improved Decision Tree Induction Based On Lookahead

摘要：

Traditional Iookahead method would not Improve the quality of decision tree. To overcome this problem, an efficient way of incorporating rough sets reduction is presented in this paper-the constrained Iookahead method. The improved algorithm restricts the search space. Instead of searching for the optimal split attribute within the whole attribute set, attribute reduction method is used to choose a subset of attributes, and then we find the optimal split attribute among the pre-selected attribute sets to construct the tree through a k-step Iookahead. Experimental results show that the proposed method decreases the model size, and improves the test accuracy.

关键词： decision tree attribute reduction constrained lookahead

作者: Yue Zhang Junhai Zhai Xizhao Wang

作者单位: Key Lab. of Machine Learning and Computational Intelligence College of Mathematics and Computer Science, Hebei University Baoding, China

会议类型: 国际会议

会议名称: The 10th International Conference on Intelligent Technologies(第十届智慧科技国际会议 InTech09)

会议地点: 桂林

会议语种:英文

页码: 671-675

在线出版日期: 2009-12-12（万方平台首次上网日期，不代表论文的发表时间）