FraudGuard: Advancing Credit Card Fraud Detection with Machine Learning Benchmarks

This research presents a comprehensive approach to credit card fraud detection, combining machine learning algorithms with user-centric applications. Analysing trends and fraudulent activities across various parameters, the study enhances fraud prevention strategies for stakeholders. Employing algorithms like XGBoost, Decision Tree, and Random Forest, and using SMOTE to balance the dataset, XGBoost achieved the highest accuracy. Practical tools developed include a real-time fraud detection app using Streamlit and a Tkinter-based desktop GUI, both leveraging the trained XGBoost model for instant predictions.