A new Bayesian classification algorithm based on attribute reduction
Hongmei Nie, Jiaqing Zhou · 2015
Naive Bayesian classifier is a simple and efficient classification method.However, the assumption of the independence of its attributes is difficult to be satisfied, which influences the classification performance.In this paper, a new classification algorithm is proposed, which is based on the attribute correlation coefficient and principal component analysis.By the algorithm, we can remove the attributes that are not related to the class, and make sure that the retained attributes are independent of each other.By removing redundant attributes, the obtained attribute subset meets the assumption of Naive Bayesian classifier, and ultimately improves the classification performance of Naive Bayesian classifier.