Loan Approval Prediction Using Machine Learning

The system of banking has so many different products from which to profit, but its major root of earning is its credit system as the credit system can profit from the interest on the loans that they credit. For numerous issues, the system of banking requires a precise model. Predicting the defaulters of credit is a very crucial task for any bank. However, by predicting the defaulters of loan, banks can surely minimize their loss by bringing down their non-profit making assets, so that the loans already approved can be recovered without any sort of depletion and it can compete as a major framework in the statement of the bank. This emphasizes the significance of researching about the forecasting of the approval of loan. The various techniques used in the Machine Learning (ML) are extremely important and critical in foreseeing this sort of data. In this work, various ML algorithms based on classification are used: Logistic Regression (LR), Decision Tree (DT), K-nearest neighbor (kNN), Multilayer Perceptron (MLP), Random Forest (RF), with the Random Forest algorithm being the most precise to predict the approval of loan with high reliability. The novelty of the technique is that it results in a highly accurate model using RF classification technique as it is based on both exploratory data analysis as well as attributes of client. Its accuracy can be found to be more than many of the existing RF approaches.