Credit Card Fraud Prediction Using XGBoost: An Ensemble Learning Approach

摘要

With the development of technology, the internet and eCommerce online payment has become an essential mode of payment. Nowadays, credit card payment is a convenient mode of payment online as well as offline transactions. As online credit card payment increases, fraud transactions are likewise increasing day by day. Increasing fraud transactions in the online payment system became a more significant challenge for banks, companies, and researchers. Therefore, it is essential to have an efficient methodology to detect fraud transactions while payment has completed via credit card. Although many traditional approaches are already available for fraud transaction prediction, existing methods lack accuracy, and it can be increased by ensemble techniques such as XGBoost. In this paper, the authors use an ensemble approach that is XGBoost (eXtreme gradient boosting) for credit card fraud prediction. The results are compared with existing machine learning approaches.