Scorecard Development

Abstract

With the continuous improvement in the banking domain many humans now are making use of bank loans but the bank has its confined resources which it can grant to specific people only, thus finding out to which person the loan may be granted such that it is a secure alternative for the bank is a complicated procedure. One of the quality indicators of the loan is loan status. It would not display the entirety at once, however it acts as a primary step of the loan lending process. So in this project we attempt to lessen this risk factor in the back of selecting the safe person in order to reduce efforts of banks and save its assets. Recovery of loans is a very crucial factor contributing in the financial statements of a bank. It is very hard to forecast the probability of loan repayment by the customer. In past years, loan approval prediction system is a domain that many researchers have worked on. Machine Learning strategies are very helpful in forecasting results for large amounts of data. In this project, we build a scorecard by carrying out various number of steps and tasks, that helps in scoring each loan applicant based on certain characteristics. That score further helps to decide a bank whether that applicant would be safe to grant a loan or not.