Credit risk modelling is at the core of any credit department in a major bank you can think of, and it revolutionizing the ways of availing facilities with tailored scores helping banks to maximize their risk return ratios and minimize the loss given default statistics.
The advances in financial technology (fintech) has facilitated new ways of doing business in the financial services industry. Financial institutions involved in lending are now compelled to push their offerings closer to the market thanbefore, be more transparent and non-discriminating in their credit underwriting, and to reduce their loan application turnaround.
Credit scoring is a statistical method which is used to predict the probability that a loan applicant, existing borrower, or counterparty will default or become delinquent. In this project, we'll be employing Machine Learning Techniques, to an existing customer dataset to determine the likelihood of default as well as generate a credit score for each client.
This will help guide a financial institution in credit risk management by informing their decision making process in whether or not to offer credit as well as credit amounts to be offered.
Typically, the pipeline involves
- Gathering tha data
- Data Quality/Data Cleaning
- Developing the behaviourial scorecard (The machine learning/statistical model)
- Using the model to predict scores (Credit rating)
- Introducing penalty factors to check for certain conditions
- Validating the model using industry specific approaches(Statistical validation)
- Python
- NumPy
- Pandas
- Matplotlib
- Scikit-learn