A LOAN APPLICATION APPROVAL SUPPORT SYSTEM USING FUZZY LOGIC MODEL

ABSTRACT

Financial Organizations always face problem of given loans to potential borrowers who request for loans. This arises as a result ofbad experience by loan defaulters. There is therefore the need for financial institutions toalways evaluate the borrowers ability to pay back loan. Several methods have been used to solve this problem which include expert judgement and machine learning algorithms. Machine learning have been found to be better than the expert judgements. This research proposes a machine learning approach to loan approvalusing fuzzy logic model in orderto address the shortcomings of the existing loan approval methods. Fuzzy logic can handle imprecise and uncertain information very well. The input variables to the systems are income, credit history, employment, criminal record, and collateral. The model was developed using React TypeScript, Golang, and MATLAB. Experimental assessment of the model gave 90% accuracy in predicting borrower creditworthiness.

KEYWORDS: Fuzzy Logic, Loan Management System, Machine Learning, Model