中文摘要：

朴素贝叶斯分类是一种简单而高效的方法,但是它的属性独立性假设,影响了它的分类性能。为了克服该问题,提出了一种基于概率推理的加权朴素贝叶斯分类模型。通过计算属性和类之间的相关概率和不相关概率,对属性赋予不同的权重,从而在保持简单性的基础上有效地提高了朴素贝叶斯算法的分类性能。实验结果表明,该方法可行而且有效。

英文摘要：

Naive Bayes Classifier is a simple and effective classification method, but its attribute independence as- sumption makes it unable to express the dependence among attributes in the real world, and affects its classification performance. In this paper a method for setting attribute weights based on probabilistic inference for using with Naive Bayes is presented. It improve the classification performance of Naive Bayes through set attributes different weights by their related - probability and unrelated - probability with classes. Experimental results illustrate the effi- ciency of this method.