Attribute discretization method based on rough set theory and information entropy

This paper analyzed the current attribute discretization methods,and put forwards a cut point set computing algorithm.Then defined the cut information entropy,and used it as the importance measurement of cut point.Based on the above method,put forward an attribute discretization algorithm based on the rough set theory and information entropy.In the experiments,detected the effectiveness of the method and compared with other methods.The results show that the method is effective when the sample and attributes number increasing to very large.