Abstract Credit institutions to provide variety of facilities to their customers, need to comprehensive studies by qualitative and quantitative aspects of their applicants. By this way, accomplish a complete evaluation of repay ability measure and calculate the refund facilities probability and finance services by them , these reviews generally validation name. The purpose of this study was ranking customer groups and specifies the best part of them until brokerage firm do its credit allocation process mechanically. Here, after the preprocessing of the data, they are processes in the RFM model. Then SOM neural network as one of the clustering algorithms will change customers to 10 cluster. Using the proposed model, the clusters will rank. The top clusters, identification and facilities grant operations to the members of these clusters will do. Finally, three clusters 5, 1 and 7 defines as top clusters that they are the target customers. Coefficient facilities granted to the top three clusters respectively are 0.271, 0.173 and 0.556.

Abstract This research has been done with the aim of identification of the effective factors that influence credit risk and designing a model for the credit rating of the legal clients of Tejarat Bank in 2003-2004 by using Data Envelopment Analysis. For this purpose, the necessary sample data on financial and nonfinancial information of 146 companies (as random simple) was selected. In this research, 27 explanatory variables (including financial and non-financial variables) were identified and examined. Finally, with the application of factor analysis and Delphi method, 8 variables, which had significant effect on credit risk, were selected and entered into the DEA model. Efficiency of the companies was calculated by using these variables. Then the model validity was measured by regression analysis. The DEA credibility scores represented the dependent variables while the 8 ratios used were considered as independent variables. The findings of the research showed that 25 companies stand on the border of efficiency. Also with one exception (owners equity/ total asset), ’all variables had the expected direction α = %5 . Research conclusions confirmed the hypothesis of DEA model’s efficiency on credit rating of the companies who have taken credit facilities from branches of Tejarat Bank in Tehran city.