A Novel Classification Algorithm based on Rough Sets and RVM Classifier
Hongyu Jia · International Journal of Digital Content Technology and its Applications · 2012
A novel classification algorithm based on rough sets and relevance vector machine classifier is proposed in this paper. Relevance vector machine is a novel support vector machine with Bayesian framewor.As redundant features can affect the classification performance of relevance vector machine, rough sets can simplify the redundant features which can improve the classification performance of relevance vector machine. In this study, we employ 180 cases to testify the superiority of the proposed classification algorithm compared with the other classification algorithms. It can be seen that RS-RVM has higher diagnostic accuracy for bearing than RS-SVM and RS-ANN.