• Version
• Download 7
• File Size 449.18 KB
• File Count 1
• Create Date 22 January 2026
• Last Updated 22 January 2026

FraudX: A Web-Based Credit Card Fraud Detection System Using Ensemble Learning and Real-Time Predictive Analytics

D. NANDHINI., MCA,

(Assistant Professor, Master of Computer Applications)

M. SUDHARSHANNAN., MCA,

Christ College of Engineering and Technology

Moolakulam, Oulgaret Municipality, Puducherry – 605010.

Abstract

The exponential rise in digital transactions has made credit card fraud a critical threat to financial institutions and consumers worldwide. Traditional rule-based fraud detection systems are increasingly ineffective against sophisticated and evolving fraudulent patterns due to their inability to adapt to new tactics and handle large-scale transaction data. This paper proposes an end-to-end FraudX that classifies transactions as legitimate or fraudulent using ensemble machine learning and real-time predictive analytics. The system utilizes historical transaction data in CSV format, applies advanced preprocessing and feature engineering techniques, and employs XGBoost as the primary classifier, alongside comparative models such as Logistic Regression and Random Forest]. Experimental results on a publicly available credit card fraud dataset demonstrate that XGBoost achieves superior performance with an accuracy of 99.8%, precision of 92.5%, and recall of 88.3%. The system is implemented as a modular web application with a Flask backend and SQLite database, offering an interactive dashboard for transaction screening, detailed reporting, and administrator-led model retraining. This work provides a scalable, efficient, and reliable solution for real-time fraud detection in financial ecosystems.

Keywords: Credit card fraud detection, ensemble learning, XGBoost, machine learning, Flask web application, financial security, predictive analytics, imbalanced data.