The Research Of F-AHP Method In Industrial Enterprise Credit Assessment

This article has made three improvements in the credit rating for industrial enterprise using F-AHP method: the index system designation, the index weight computation and the membership degree definition.In designing the hierarchical index structure, we obey the rule strictly, trying to be actual, comprehensive, scientific, comparable and operable .As the research object is special, we design a whole set of index system for industrial enterprise only. At the same time, we divide the index system into four levels, it principally includes four classes, those are firm’s basic quality, financial status, historical credit register and develop foreground. The index set contains 35 indices,including 25 qualitative targets and 10 quantitative targets.When we compute the weight of indexes, we use the AHP method. At first, we design judgment matrixes by using scaling methods.Then we checkout the consistency of those matrixes. And at last, we compute the weights for indexes and calculate the integrative weight. Through each detail step, we balance through methods, and find out the best means which match with our index system. We contrast scaling methods, and find out that 0.5-0.9 scaling method suits better to our index system than 1-9 scaling method, so we choose the former to design judgment matrixes; And through balancing weight calculating methods, we find out that the computing results of the Sum Product Method and Root Method are the same.In ascertaining the membership degree, we use the Delphi law to account the membership degree for qualitative indices. We invite professional experts to assess the qualitative indices of the industrial enterprises. To certain the membership degree of quantitative targets, we use the sectionalized membership grade function. By classifying the quantitative targets into three kinds: the intermediate type, the efficient type and the cost type, we design membership function for them separately. This kind of function efficiently solute the problem that comment jump obviously from one to another when slight turbulent motion occurs to the target value around the critical value.And then compose the fuzzy and AHP methods to evaluate comprehensively.