混合树增广朴素贝叶斯分类模型

Mixed tree augmented Naive Bayes classifier model

树增广朴素贝叶斯分类算法(TANC)虽然降低了朴素贝叶斯分类算法(NBC)的条件独立性约束,但是该模型同时又要求每个条件属性结点(除树的根结点外)都有两个父结点,这种限制同样降低了分类的正确率。因此,提出了一种基于粗糙集理论的混合树增广朴素贝叶斯分类模型(MTANC)。通过在UCI数据集上的仿真实验,验证了该方法的有效性。

Although tree augmented Naive Bayes classifier（TANC） model reduces the independence assumption restriction of Naive Bayesian classifier（NBC） model,the model requires each attribute note（except for the root note） to own two parents,which has decreased the accuracy rate of classification.So a new Bayesian model mixed tree augmented Naive Bayes classifier（MTANC） based on the rough set theory is presented.The validity of the approach is established by the simulated experiment of UCI data set.