Indian Journal of Science and Technology

Zahra Mahdianfar^*

Background/Objectives: Today credit risk is recognized as one of the biggest causes of bankrupting of banks and financial institutes. To control and manage this risk, designing credit rating models is an undeniable necessity.
Methods/Statistical Analysis: Research method due to providing artificial intelligence concepts including neural network and fuzzy inference system and also due to the probability of its implementation in banking system, regarding to purpose it is developmental applicable and in terms of implementation, it is descriptive-analytic. In this research, we have introduced an overall model of credit rating and its database. Results: Since a high percentage of bank customers are natural persons, attention to the effective characteristics of customers in credit rating is very important. This paper has investigated the scale of the descriptive and qualitative effects (indecisive) of the customers such as commitment, faithfulness, etc. in credit rating. Although there have been proposed various methods and models based upon customers` characteristics, yet there has never been any critique and investigation regarding the scale of effects of indecisive features. The main objective of this paper is to investigate the scale of effects of qualitative effects in their credit rating and comparing their qualitative features with other quantitative features of customers using Neural Network and Fuzzy Inference System, in a way that it makes determining the measurements of customers’ credits easy for the credit providers.
Conclusion/Application: This research indicates that we cannot assess the customers only by relying on decisive, clarified indexes and their financial status, and the status of indecisive variables effectiveness (oral variables).

Keywords: Credit Rating, Customer Qualitative Feature JEL:C53 C40, Fuzzy Inference System, Neural Network