Skip to main content
Fig. 6 | Financial Innovation

Fig. 6

From: A dynamic credit risk assessment model with data mining techniques: evidence from Iranian banks

Fig. 6

Conceptual model of the dynamic model. T1, t2 … tn show the months. We organize a table based on the monthly data on bad customers (i.e. those who have not paid their debt for the last 2 months or so), then we apply FCM to cluster customers, then we train the ANFIS and construct a dynamic model based on the data in this table. This model is used then to assess the customers’ credit risk at the time of registry. Only if the customer is assessed to be low risk, they are given credit. Otherwise, if they are found high risk they will get no credit. Alternatively, if the customer belonged to the medium-risk segment in the analysis with the dynamic model, the second round of assessment will begin using a fuzzy inference system based on our predefined rules. The analysis ends if the customer is still shown to be too risky. However, if they are shown to belong to the medium risk group, conditional credit can be allocated to these customers

Back to article page