互信息匹配的半朴素贝叶斯分类器.pdf

84 2016 ，52（18） Computer Engineering and Applications 计算机工程与应用 互信息匹配的半朴素贝叶斯分类器 1 2 2 赵 亮 ，刘建辉 ，崔彩峰 1 2 2 ZHAO Liang , LIU Jianhui , CUI Caifeng 1.辽宁工程技术大学 研究生院，辽宁 阜新 123000 2.辽宁工程技术大学 电子与信息工程学院，辽宁 葫芦岛 125000 1.Institute of Graduate, Liaoning Technical University, Fuxin, Liaoning 123000, China 2.School of Electronic and Information Engineering, Liaoning Technical University, Huludao, Liaoning 125000, China ZHAO Liang, LIU Jianhui, CUI Caifeng. Semi-naive Bayesian classifier matched by mutual information. Computer Engineering and Applications, 2016, 52 （18）：84-87. Abstract ：Because the class-conditional independence assumption which is the mainly feature of Naive Bayesian classifier is a so strong assumption and difference appears between datasets, the class-conditional independence assumption becomes the entry point of improvement methods. But some researches indicate that the violations of independence assumption do not make so much influence to the classifier as expected. This paper proposes a conditional entropy matching half-naive Bayesian classifier for the purpose of lower posterior probability estimation error. Experiments show that this method can effectively improve the performance of naive Bayesian classifier. Key words ：semi-naive Bayesian classifier; mutual information; matching 摘 要：由于作为朴素贝叶斯分类器的主要特征的条件独立性假设条件过强且在不同数据集上表现出的差异，所以 独立性假设成为众多改进算法的切入点。但也有研究指出不满足该假设并没有对分类器造成预想的影响。从降低 后验概率的估计误差入手提出一种条件熵匹配的半朴素贝叶斯分类器。实验证明，该方法能有效提高朴素贝叶斯 分类器的性能。 关键词：半朴素贝叶斯分类器；互信息；匹配 文献标志码：A 中图分类号：TN911.7 doi ：10.3778/j.issn. 1002-8331.1412-0318 1 概述 Lazy Bayesian Rules[3] 和 TAN （Tree-Augmented Naive [4-5] 朴素贝叶斯分类器是机器学习领域最著名的分类算 Bayes） 。另一类则以相关性为依据对属性赋权，放大与 法之一，拥有简单的运算和优秀的性能，虽然提出时间