Fuzzy cluster in credit scoring

Nine companies listed on China Stock Exchange by 2000 are chosen and the following six major financial indexes of them are considered: net assets yield, net profit per stock, receivables velocity, stock velocity, floating ratio and asset/debt ratio. Using fuzzy dynamic cluster analysis, this paper classifies these 9 listed companies into three types: Good, Middle and Bad, then two most important financial indexes in direct ratio to the financial status: net assets yield and receivables velocity are identified. They are abstracted into a subject function representing this type through trapezium distribution. In doing so, a fuzzy cluster evaluation standard is established. Finally, by comparing the listed companies being scored with the fuzzy cluster evaluation standard, and according to the maximum subject principle, the credit scoring for the companies can be obtained.