Fraud Detection In Credit Card Transaction By Machine Learning

Abstract

Systems can cluster data and provide incredibly accurate outcomes under machine learning. Using the
XGBoost algorithm, this study investigates machine learning for fraud detection in an effort to improve corporate
procedures and lower fraudulent activity in big businesses. We suggest an XGBoost-based model for online credit
card fraud detection that has been validated through case studies from two banks. This model is intended for realtime
fraud detection. The results show how well XGBoost detects fraud while striking a good balance between recall
and precision, greatly increasing the effectiveness of financial systems. Analysis using Python demonstrates how
machine learning models can handle and stop fraud on dynamic datasets in real time. This study concludes by
demonstrating that machine learning algorithms such as XGBoost may be used to dynamically manage fraudulent
actions, efficiently handle online credit card fraud detection in banks, and continually improve the fraud detection
and prevention system..