It simply introduces the basic thought of Bayesian decision tree method in this paper,which takes advantage of the prior infor-mation method for Bayesian classification and the information gain method of decision tree,and makes up for the decision tree cannot handle the ambiguity data and the missing value by adding Bayesian node. On this basis, a Bayesian decision tree algorithm based on Naïve Bayesian method and ID3 algorithm is presented named NBDT-ID3 algorithm. The algorithm process of the design and analysis is introduced. Then the algorithm is applied to higher vocational admission data mining,which analyzes and forecasts the new student regis-tration. It is tested and verified under the Matlab environment. The experimental results show that NBDT-ID3 algorithm not only can get higher classification accuracy but also behave well in handling the ambiguity,incomplete or incongruous data in the case of paying certain of time.%文中首先简单介绍了贝叶斯决策树方法的基本思想,该方法结合了贝叶斯分类的先验信息方法和决策树分类的信息增益方法的优点,加入贝叶斯节点弥补了决策树不能处理具有二义性或存在缺失值数据的缺点。在此基础上,文中设计了一种基于朴素贝叶斯方法和ID3算法的贝叶斯决策树算法———NBDT-ID3算法,并给出了该算法的设计及分析过程。然后将该算法应用到高职招生数据挖掘中,对新生报到情况进行分析与预测,并在Matlab环境下进行了实验验证。实验结果表明,NBDT-ID3算法在付出一定时间代价的情况下,不仅可以获得更高的分类精度,而且在处理二义性、不完整或不一致数据方面具有更好的效果。
展开▼
