Fuzzy Expert System for Bank Loan Approval

Background: Bank loan approval is one of the important pillars of the banking system; it is the process of approving or denying a loan to companies or individual customers by the bank. The approval process has a lot of parameters to be taken into consideration, which is ambiguous in nature; hence, bank loan approval required special knowledge to be executed. Aim: Due to the ambiguity of the approval process, we proposed the use of a fuzzy expert system which proved to handle such ambiguous problems to help banks easily and accurately make decisions on bank credit approval. This proposed fuzzy expert system will help banks in making accurate decisions easily even in the absence of the domain expert on credit approval based on the knowledge of an expert in the field. Method: The proposed fuzzy expert system was developed using a fuzzy tool in MATLAB software and it has two stages, where the first stage decides on three output parameters which are repayment, ability manage, and risk. Total asset, credit repayment, 18% earning, business stability, credit missed, asset/debt ratio, bond rating, and dollar to Naira ratio are the input parameters for the first stage of the system. The second stage of the system used the output parameters values of the first stage as its input parameters to make the final decision on whether to approve the credit or not. Results: Using 0.938, 0.583, 0.715, 0.88, 0.104, 0.897, 0.842, and 0.856 membership degree for total asset, credit request (loan amount), 18% earning, business stability, asset/debt ratio, bond rating, and dollar to naira ratio respectively as an input to the first stage of the system, the resultant output were 0.625, 0.367, and 0.25 for repayment, ability manage, and risk respectively, and those were feed to the second stage and result in 0.656 loan membership degree which means the loan can be approve to the customer