Loan eligibility prediction using machine learning algorithm

Lee, Zhen-Hong (2024) Loan eligibility prediction using machine learning algorithm. Final Year Project, UTAR.

Preview

PDF Download (17Mb) | Preview

Abstract

Today banks have become a principal instrument for offering both physical persons and organizations with the necessary financial means banks require for such goals like property acquisition or project financing. What ultimately decides this is the contemplation of a borrower's creditworthiness and chances that he or she will repay back what was borrowed. The implementing of a loan qualifier prediction system brings enormous advantage to the lenders, banks, and financial institutions. This helps in reducing the gap between the two phase that is loan application process and its decision- making process where credit is extended to the appropriate applicant based on their risk level. The project objective is to create and analyse a comparative model illustrating how to employ different machine learning algorithms in domains like loan approval processes, pattern recognition, limitations assessment, and performance metric evaluation. The study incorporates three prominent machine learning algorithms: Predicting the target variable using logistic regression, decision tree and it’s variant random forest for credit scoring model. The analysis's findings prove that, in terms of both accuracy and error, the RF algorithm is the best out of three models. The final product of the project is loan eligibility prediction website with machine learning model implemented for real life scenario use.

Actions (login required)