Prediction of the Approval of Bank Loans Using Various Machine Learning Algorithms

Both individuals and businesses now frequently obtain loans from banks, but because banks have limited resources, they carefully consider each applicant before approving a loan. To determine who the loan can be granted to and who would be a better choice for the bank, for this reason, credit institutions go through a protracted phase of verification and identification. This study aims to employ various machine learning techniques to analyze enormous amounts of data in order to detect patterns that could be used to predict whether a loan would be approved for a particular person or organization and determine whether the borrower can afford to repay the loan or not. Another goal of this study is demonstrating how machine learning will give the banking sector a more reliable predictive modeling tool than it currently has. In this work, the data was gathered from 1000 records of applicants with 16 features from the reputed bank in the city of Dhaka, Bangladesh. Finally, the following seven data mining techniques are employed for accuracy, and suggestions are made regarding how to make the loan approval process more effective and affordable for both banks and borrowers by dramatically lowering the time and effort needed: Decision Tree, Gradient Boosting, Random Forest, Naive Bayes, Support Vector Machine, and Logistic Regression are some examples of machine learning techniques. At the conclusion of the study, the Random Forrest method has the highest accuracy.