Granular mappings.pdfVIP

292 IEEE TRANSACTIONS ON SYSTEMS, MAN, AND CYBERNETICS—PART A: SYSTEMS AND HUMANS, VOL. 35, NO. 2, MARCH 2005 Correspondence________________________________________________________________________ Granular Mappings Andrzej Bargiela and Witold Pedrycz Abstract—In this paper, we are concerned with the granular representa- tion of mappings (or experimental data) coming in the form : [0 1] (for one-dimensional cases) and : [0 1] (for multi- variable cases) with being a set of real numbers. As the name implies, a granular mapping is defined over information granules and maps them into a collection of granules expressed in some output space. The design of the granular mapping is discussed in the case of set and fuzzy set-based granulation. The proposed development is regarded as a two-phase process that comprises: 1) a definition of an interaction between information gran- ules and experimental evidence or existing numeric mapping and 2) the use of these measures of interaction in building an explicit expression for the granular mapping. We show how to develop information granules in case of multidimensional numeric data by resorting to fuzzy clustering (fuzzy C-means). Experimental results serve as an illustration of the proposed ap- proach. Index Terms—Fuzzy clustering, fuzzy relational equations, granular computing, information granules, possibility and necessity measures, specificity of granular information. I. INTRODUCTION AND PROBLEM STATEMENT The notion of information granularity and granular computing [1], [15]–[18] has emerged as one of the paradigms of computing that per- meates a broad array of pursuits of computational intelligence and con- tributes as one of its fundamentals. With the notion of information granules and granularity being central to various aspects of problem abstraction and modularization, we encounter different formal ways of expressing information granularity. The notion of granularity is formal- ized in the language of sets (quite often all