Mobile Application Development for Bank Loan Approval Prediction System using Machine Learning

Authors

  • Dr. A. Mary Francina, Dr. S. Naga Poornima, Ms. K. Mary Keerthi, Ms. Nirmala Priya D.

Abstract

Taking out a loan from a financial organization has become a widespread practice in today's environment. The number of people applying for loans is increasing every day, and they do so for a variety of different reasons. The banking system, which has a diverse range of products to provide, relies heavily on the credit system as its principal source of income. Because the interest earned on the loans, source of money. A big number of people, on the other hand, are asking for loans. It is not necessary that all application has been trustworthy, and the applicants can able to pay back the loan. Every year, we learn a batch of instances in which borrowers fail to repay their loans in full, resulting in considerable losses for the lending institutions. It's difficult to find a faithful borrower, he or she would repay the loan. When a procedure is carried out manually, there areplenty of possibilities for human mistakes. Predicting which customers are true is one of the most difficult tasks for any financial institution. This is very much important for further investigation of the loan approval prediction. To automate the process of granting bank loans, banking systems require an accurate modelling system to be implemented. To address this issue, aMachine Learning (ML)algorithm such as Logistic Regression (LR), Random Forest (RF), and C4.5 was applied on collected Kaggle data. The metrics of LR, RF, and C4.5 are used in a comparison to determine which ML algorithm is the most appropriate for the task at hand. The RF model was selected because of its high accuracy rate of 90.86%. A smartphone application has been developed to assist clients in saving both time and travel. Customers can use the app to determine whether or not they are eligible for a loan.

Downloads

Published

2022-04-28