决策树算法的研究及优化

决策树算法是数据挖掘中的一个比较活跃的研究领域,是对分类问题进行深入分析的一种方法,但构造最优决策树是一个NP困难问题。文中首先介绍了ID3算法的基本思想,然后讨论了决策树算法中的难点问题,针对ID3算法中所存在的不足,提出了一种利用优化法的思想来改进信息增益的算法,并且与ID3算法进行了实验对比。通过实验表明,这种方法从树的规模和分类精度都优于许多决策树算法,使决策效率明显提高。

Research and Optimization of Decision Tree Algorithm

Decision tree is one of heated fields in data mning in recent years,and it is a widely-used solution for classification problems. But the design of the optimal decision tree has been proved to be NP-hard. This paper first introduces the main thoughts of algorithm of ID3, then discusses some difficult problems of selecting value on division in decision tree,and puts forward an algorithm to improve the gain of information, and makes a contrast to ID3 by experiments. Results demonstrated this algorithm outdoes the others in the size of decision tree and classification accuracy.