Comparative Analysis of Bank Loan Defaulter Prediction Using Machine Learning Techniques

Abstract

Nowadays, there are numerous risks identified with the banking sector regarding giving loans to the clients and for the individuals who get the loan. The examination of risk in bank credits needs to understand what is the reason for this risk. Likewise, the quantity of exchanges in the financial area is quickly developing and information volumes are accessible which address the client’s conduct, and the risk of giving loans are expanded. The objective of this paper is to discover the nature or details of the clients who are applying for the loan. This paper proposes a comparative study of three machine learning models, namely, Random Forest, Naive Bayes (Gaussian model, Multinomial model, and Bernoulli Model), and Support Vector Machine (Linear kernel, Gaussian RBF kernel, and Polynomial kernel), to predict whether a customer may get a loan or not. In this paper, we analyze the evaluation parameters, namely, classification accuracy, precision, recall, and F1-Score for these machine learning models to foresee which model is best suitable for predicting a loan.