Research on the Decision Tree Classification Algorithm and an Application in Estate Evaluation
In this paper, we introduce the algorithm of C4.5 which is a decision tree algorithm and two improving technologies of C5.0: boosting and the technology of cost-sensitive tree constructed by instanceweighting method. We use the decision tree classification method based on C5.0 to evaluate estate. The results show that it is very evident to increase the classification accuracy by using boosting. After using boosting, the classification accuracy rate rises from 97.1％ to 98.6％ in the training set and from 89.1％ to 92.9％ in the testing set. Furthermore, we can reduce the misclassification number to giving a higher cost value so that the users requirements can be satisfied under different situations. All in all, the method of decision tree based on C5.0 is very effective and practicable.
Jizhang Gong Sulin Pang Gang Hao Jian Ma
Department of Accountancy and Mathematics and Institute of Finance Engineering, Jinan University Gua Department of Management Sciences City University of Hong Kong Kowloon Tong, Hong Kong, China, 99907 Department of Information Systems City University of Hong Kong Kowloon Tong, Hong Kong, China, 99907
国际会议
南宁
英文
2007-07-20（万方平台首次上网日期，不代表论文的发表时间）