三种典型贝叶斯分类器的研究
首发时间:2009-10-19
摘要:贝叶斯分类方法是数据挖掘中一种重要的分类算法。在贝叶斯家族中有三种典型的贝叶斯分类器:朴素贝叶斯分类器、TAN贝叶斯分类器和贝叶斯网络分类器。本文主要研究了TAN分类器中根结点的设置对分类影响,以及将这三种典型贝叶斯分类器应用到5个典型UCI数据集上,分析比较它们对不同类型和规模数据集的分类情况,总结这三种分类器的适用范围。
关键词: 朴素贝叶斯分类器 TAN贝叶斯分类器 贝叶斯网络分类器 根结点 UCI
For information in English, please click here
Research on Three Typical Bayesian Classifiers
Abstract:Bayesian classifer is one of the important classification algorithms in data mining. There are three kinds of typical bayesian classifiers in the bayesian family: naive bayesian classifier, TAN bayesian classifer and bayesian network classifier. At first, this paper mainly studies the impacts of the setting of root nodes on classification results of TAN bayesian classifiers. Secondly, we apply these three bayesian classifiers to five typical UCI datasets and analyze the classifying situation on UCI datasets with different types and size. At last, conclusion of the application field of the three kinds of bayesian classifiers is given.
Keywords: Naive bayesian classifier TAN bayesian classifer bayesian network classifier root node UCI
基金:
论文图表:
引用
No.3591449647812559****
同行评议
共计0人参与
勘误表
三种典型贝叶斯分类器的研究
评论
全部评论0/1000