Credit Sanction Forecasting

Abstract

In today's banking sector, there is a lot of enhancement because of the advancement of technology. The number of applicants for the loan approval is also increasing every day, and it is difficult for the banking sector to verify each applicant manually and then recommend for the loan approval. The banking sector still needs a more precise method for forecasting the safe customer before approving the loan. One of the quality metrics of the loan is the status of the loan. It doesn't instantly reveal anything, though it is a foremost step in the process of loan approval. To obtain a defaulter and also valid user, the Credit Sanction Forecasting framework is used for precise analysis of the credit data. A customer's loan repayment capacity is more reliably estimated using the random forest classifier technique. Therefore, the efficiency of this projection is based on the multiple factors of the Random Forest method. The aim is to show that parameter optimization outcomes in high accuracy for the estimation of loan repayments capacity by customers. The primary aim has implemented using a software package of python and machine learning algorithms. The combination Min-Max standardization, Logistic Regression, Random Forest classifier, and deep learning model created using tensor flow are used to predict the safe customers for the loan approval. CSF offers important details with high accuracy and is also mainly used to forecast the loan status of the bank with help of a classification algorithm of ML and deep learning.