Abstract: The discretization of continues attributes is always with great contribution to the followed process of machine learning or data mining. A new algorithm based on information entropy for discretization of decision table was proposed. Through inconsistency checking of decision table, we deleted some redundant cut points on the basis of preliminary discretization scheme. The experiments of classification of discreted data were performed by using SVM, and meanwhile compared with other algorithms, the presented algorithm is effective.

Key words: Discretization,Decision Lable,Information entropy,Inconsistency