Rough-fuzzy Classifier Modeling Using Data Repository Sets

This paper reflects the trends of the past years based on the diffusion of various traditional approaches and methods when tackling new problems. Two components of the computational intelligence (CI) are applied, rough and fuzzy sets theory. These components permit one to operate with uncertainty data. The current knowledge in the investigated field is summarized and briefly explained. It also deals with uncertainty in an information system and the two approaches, the fuzzy sets (FSs) and rough sets theory (RST), for operating it. The proposal and implementation of a rough-fuzzy classifier (RFC) is modified. RFC uses the rules generated by RSTbox. The databases IRIS and WINE were chosen for verification. The classification results were compared with the results of other classification methods are applied on these databases. Finally, we summarized the presented problems. Based on the above stated facts it can be claimed that the proposed modified algorithm, RSTbox and RFC model are functional. The model is relatively successful (compared to other approaches), and by using it two classification databases can be carried out. This model is proposed in MATLAB.