：朴素贝叶斯分类器是一种简单而高效的分类器，但是它的属性独立性假设使其无法表示现实世界属性之间的依赖关

系，以及它的被动学习策略，影响了它的分类性能。本文从不同的角度出发，讨论并分析了三种改进朴素贝叶斯分类

性能的方法。为进一步的研究打下坚实的基础。-: Naive Bayesian classifier is a simple and efficient classifier, but its attribute independence assumption that the real world so that it can not be said to rely on the relationship between attributes, as well as its passive learning strategies, the impact of the classification of its performance . This article from a different point of view, discussion and analysis of three to improve the performance of Naive Bayes classification method. For further research and lay a solid foundation.