Enhancing Credit Card Fraud Detection Using Cutting-Edge Advanced Machine Learning Techniques

Credit card fraud remains a major challenge in digital payments, necessitating advanced detection techniques. This study evaluates Deep Belief Networks (DBN), Isolation Forest, K-Nearest Neighbors (KNN), and Artificial Neural Networks (ANN) for fraud detection. Performance was assessed using Accuracy, Precision, Recall, and F1-Score. DBN, KNN, and ANN achieved 99% accuracy with high precision and recall (0.99), while Isolation Forest underperformed at 51% accuracy. To address class imbalance, SMOTE was applied, significantly improving model performance. Our comparative analysis highlights that the proposed models outperform existing approaches, demonstrating superior fraud detection capabilities. The findings contribute to minimizing financial losses, enhancing customer trust, and strengthening fraud prevention systems. Future work will explore real-time implementation and model resilience against adversarial attacks to ensure robustness in practical applications.