Loan Approval Prediction System using Logistic Regression and CIBIL Score

Abstract

Many individuals apply for bank loans. But the banks have limited assets, so it can grant credit to a limited number of customers. The credit gained by the customers can be a growing asset for a bank due to the earnings from interests or a liability if the customer is unable to pay the loan. A huge amount of capital that is disbursed may turn into bad debt just because the bank was not well informed about the repayment capabilities of its customer. Determining beforehand that which customer can repay the loan will be a safer option for the bank. The process of predicting whether a loan should be approved or not, can be done by bank officials by inspecting various parameters of a customer. Doing so will require manpower and capital as human employees would perform the job of prediction. To tackle this situation, there is a need for automation. Previous research in this area has shown that there are numerous strategies for reducing the number of loan defaults. However, accurate prediction is critical for profit maximisation. The proposed loan approval prediction system is a web application based on machine learning, designed to provide instant loan approval predictions to users. The application uses logistic regression to predict the probability of loan approval and also computes a credit score which is referred as CIBIL score. Overall, the loan approval prediction system is a powerful tool for individuals and financial institutions looking to quickly assess loan applications and make informed decisions. It leverages the power of machine learning to provide accurate and reliable predictions, and also provides an easy and a convenient way for users to access this functionality.