Some Experiments to Compare Rough Sets Theory and Ordinal Statistical Methods

Abstract

In statistics, classical discriminant analysis methods explain a predefined classification of a set of objects, by using the discrete values taken by these objects on some conditional attributes. Rough sets theory has the same objective. The purpose of the paper is to describe some experiments made to compare both approaches. In a first section, we recall the results obtained by Wong et al. for the comparison of rough sets theory with the Quinlan’s method, using the notion of entropy. Section two is devoted to a comparison of rough sets approach with a specific discriminant method, the Elysee method; an application is solved by both methods and similar results are obtained. In the last section, we analyse a real case study concerning a production problem, in a printing company, by both methodologies.