Baseline Modeling for Early Prediction of Loan Approval System

Loans are important aspect of any banks. Any loan application's approval or denial directly affects the bank's income and profitability. Many people come to bank for availing loans and banks have limited funds in their accounts. To approve the loan in this situation, the proper applicant must be chosen. This will make a profitable situation to both banks and customers from which both of them will be benefited and reduce the risk of loss of money. Therefore in order to make an effective loan approval many machine learning prediction models are used in our work. In order to overcome the problem of loan going into default by identifying whether the customer is eligible for loan and to help customers keep a good credit score by avoiding the loans which they cannot repay. The prediction is performed by data pre processing techniques to clean the dataset and give accurate data for training the machine learning models. Many machine learning models which are mainly used for classification algorithm are trained and are tested in our work for predicting whether to approve a loan applicant or not. Baseline modeling and KFold cross validation techniques are used to check the accuracy of models and the model with highest accuracy is taken for consideration of developing the loan approval prediction model. This trained model is then turned to receive the inputs from the user and give the predictions results as the output.