Banks generate revenue primarily by lending money to borrowers. To maximize profits and maintain a good reputation, they need to identify borrowers who may have trouble repaying their loans. When borrowers fail to make payments, we often categorize them as "delinquent" or "defaulted."
- Delinquent borrowers are slightly behind on their payments.
- Defaulted borrowers have been behind for a long time and are unlikely to repay.
- This project aims to identify borrowers likely to default in the next two years after being seriously delinquent for more than three months.
- We use various borrower characteristics and historical data to predict this risk.
- These predictions help banks take proactive measures.
Build a model using borrower information and historical records to predict serious delinquency within the next two years.
- Language:
Python - Libraries:
Pandas,Matplotlib,NumPy,Scikit Learn,ImblearnShapandLIMEfor model interpretation.Keras