Rough Set Based Fuzzy Modeling by Occupancy Degree and Optimal Partition of Projection
The rough set theory suggested by Pawlak has a property that it can represent the degree of consistency between condition and decision attributes of data pairs which don’t have linguistic information. In this paper, by using this ability of rough set theory, we define a measure called occupancy degree which can represent a consistency degree of premise and consequent variables in fuzzy rules describing experimental data pairs. We also propose a method by which we partition the projected data on input space and find an optimal fuzzy rule table and membership functions of input and output variables from data without preliminary linguistic information.
KeywordsMembership Function Fuzzy Rule Optimal Partition Fuzzy Partition Consequent Variable
Unable to display preview. Download preview PDF.
- 1.Bezdek, J.C.: Some Recent Applications of Fuzzy C-means in Pattern Recognition and Image Processing. In: IEEE Workshop on Lang. Autom., pp. 247–252 (1983)Google Scholar
- 2.Delgano, M., Gomez-Skarmeta, A.F., Martin, F.: A Fuzzy Clustering-Based Rapid Prototyping for Fuzzy Rule-Based Modeling. IEEE Trans. on Fuzzy Systems 5(2) (May 1997)Google Scholar
- 6.Inoue, H., Kamei, K., Inoue, K.: Auto-Generation of Fuzzy Production Rules Using Hyper Elliptic Cone Membership Function by Genetic Algorithm. In: Proc. of Intern. Conf. on Soft Computing, Japan, pp. 82–85 (1996)Google Scholar
- 7.Pawlak, Z.: Why Rough Sets? In: Proc. of IEEE Int. Conf. Fuzzy Systems, pp. 738–743 (1996)Google Scholar