Discretization and grouping: Preprocessing steps for data mining

Abstract

Unlike on-line discretization performed by a number of machine learning (ML) algorithms for building decision trees or decision rules, we propose off-line algorithms for discretizing numerical attributes and grouping values of nominal attributes. The number of resulting intervals obtained by discretization depends only on the data; the number of groups corresponds to the number of classes. Since both discretization and grouping is done with respect to the goal classes, the algorithms are suitable only for classification/prediction tasks. As a side effect of the off-line processing, the number of objects in the datasets and number of attributes may be reduced. It should be also mentioned that although the original idea of the discretization procedure is proposed to the KEx system, the algorithms show good performance together with other machine learning algorithms.