Discretization of Continuous Interval-Valued Attributes in Rough Set Theory and its Application

Abstract

Rough set theory is a relatively new soft computing tool to deal with vagueness and uncertainty, and is regarded as a field of leading edge. But it cannot deal with continuous attributes, thus discretization is a problem which we cannot neglect. Discretization based on rough set has some particular characteristics, and consistency must be satisfied for discretization of decision systems. Existing discretization methods cannot well process continuous interval-valued attributes in rough set theory. A new approach is proposed to discretize continuous interval-valued attributes in this paper, which enhances the precision of classification and accurate recognition rate in pattern recognition. In the simulation experiment, the decision table was composed of 3 features and 17 radar emitter signals, and the recognition results obtained from this discretization algorithm show that the proposed approach is valid and feasible. The approach expands the application scope of rough set theory.