A New Method of Rule Extraction from Quantitative Data Using the Theory of Rough Sets and Fuzzy Logic

The extraction of rules from unclassified quantitative data is one of the works done in the fuzzy-rough logic. When the rough and fuzzy algorithms are blended, the quantitative data will be shown as fuzzy variables. Then, the rules are extracted from the approximation from below and above. The investigation of the membership functions of each feature in this process is necessary and time-consuming. Monotonic membership function will be introduced in this paper. The investigation of this monotonic function is substituted for the investigation of the membership functions of the information system features. The proposed method is a simple and efficient method.