Learning Naive Bayes Classifiers with Incomplete Data

Naive Bayes Classifiers have been known with the advantages of high efficiency and good classification accuracy and they have been widely used in many domains. However, the classifiers need complete data. And the phenomenon of missing data widely exists in practice. Facing this instance, learning naive Bayes classifier and classification method with missing data are built in this paper. Compared with the common methods dealing with missing data, this method is more efficient and reliable.