The fuzzy inference rule extraction and attribute reduction based on AFS theory and closeness degrees

The attributes in database are very important information. However, sometimes they cannot be applied efficiently. What we concern about is how to make use of them as much as possible. AFS theory is a new analytic method of fuzzy mathematics. The membership functions of AFS theory are obtained by a uniform arithmetic according to the original data. AFS theory and closeness degree functions are combined to analyze the relation of the attributes in database. Using AFS theory and fuzzy inference methods, a new fuzzy inference method is proposed. At the same time, we get a method of attribute reduction. The results are very similar to human intuition and show that the methods are convenient for data mining and attributes analyzing.