A new clustering classification approach based on FCR

A new clustering classification approach based on fuzzy closeness relationship (FCR) is studied in this paper. As we know, fuzzy clustering classification is one of important and valid methods to knowledge discovery. One of problems in fuzzy clustering classification is to determine a certain fuzzy sample classification in given limited sample space. Another is its validity, that is to say, if the sample is resemble in sample space, its fuzzy type will be resemble too. In our research, firstly, using triangle arithmetic operator and triangle transference, we extend fuzzy equivalence relationship to fuzzy closeness relationship, and cluster. Secondly, importing fuzzy coverage based on fuzzy closeness relationship, we judge resemble type of resemble sample. Thirdly, we introduce the clustering classification method based on fuzzy closeness relationship and fuzzy coverage. The method can overcome information more loss in fuzzy equivalence. Finally, we test its feasibility. The approach is applied to knowledge discovery for risk prediction of electronic commerce market, good result has been gotten.