Identifying Credit Card Defaulters and Predicting Fraudulent Transactions using Various Machine Learning Techniques

In recent years as increase in number of credit card transactions, large amount of data generated via Razor pay, Billing desk payment gateways. Indian government encourages to each and every individual to do digital transactions via online payment gateway under the scheme of digital India. For going cashless economy under the banking domain credit card transactions for the booking flight tickets, online shopping, railway ticket booking has increased rapidly after the covid pandemic. Credit card banks giving offers like some amount of percentage of cashback and some credit points for the refilling the fuel on the selected petrol pumps like BPCL and HP so a greater number of users to use credit card on petrol pumps for filling up the fuel tank. This is the positive side of the digital India concept. Due to increasing number of online transactions via credit card it also increases the defaulters and fraud by using credit cards transactions. During online payment via credit card a large amount of data is generated including customer information, credit card number, what type of product customer like to purchase, credit history, transactions history and has the customer pay the credit card bill before due date. As the customers CIBIL score is generated on the basis of due date payment of credit card which is used for future reference. Credit card holder have serious implication if credit card holder is a defaulter. So, we are going to propose a solution of generating a model by using different machine learning algorithm techniques to avoid such situation. We are doing comparison of different performance models to identify defaulters using evaluation matrices such as accuracy, precision, recall and F1-score. Including decision making on historical credit card transaction using machine learning algorithms Support vector Machine, Logistic regression and Random Forest to identify credit card defaulters to preventing the financial loss of the credit card lending banks.