Rough Sets and Applications for Data Analysis

Abstract

Rough set is a framework for uncertainty (vagueness) derived from indiscernibility relation. By this idea of rough set for uncertainty, several data analysis methods have been developed, e.g. attribute reduction, rule induction, clustering and so on. Moreover, by replacing the indiscernibility relation with similarity or dominance relation, the rough-set approach can be adopted for several objectives of data analysis, e.g. analyzing data without complete information, learning preference information and so on. In this paper, studies of rough-set based data analysis, especially attribute reduction, are introduced. Moreover, several extensions of rough set model are provided.