Research and Resolve Over-fitting of Decision Tree Algorithm
Wang Qin-zhu · 2011
Decision tree algorithms are widely used in the field of pattern recognition and machine learning,and used to solve problems related with the classification.The result of classification will be largely affected by over-fitting problem in the decision tree algorithm.According to the reasons of over-fitting,this paper analysis student achivement decision data by pessimistic error pruning method and get some important factors that affect students' performance.Experiments show that the method can get a classification rules as short as possible,effectively improve the performance of the decision tree.