Generative AI-Based Data Management and Predictive Analytics for Financial Fraud Detection in Banking Systems

Abstract

Financial fraud in banking systems has become increasingly sophisticated with the rapid growth of digital transactions. Traditional fraud detection techniques often struggle to identify complex and evolving fraud patterns in real time. This project proposes a Generative AI-based data management and predictive analytics framework for effective financial fraud detection in banking systems. The system leverages Generative AI to simulate realistic transaction data, improving model training and handling data imbalance issues. Advanced machine learning algorithms such as Random Forest, Support Vector Machine, and hybrid models are used to analyze transaction behavior and detect anomalies.The proposed framework integrates efficient data preprocessing, feature extraction, and predictive modeling to enhance detection accuracy and reduce false positives. By combining generative and predictive approaches, the system adapts to new fraud patterns and ensures scalability in large banking environments. Experimental results demonstrate improved performance in terms of accuracy, precision, recall, and F1-score compared to traditional methods. This approach provides a reliable, intelligent, and scalable solution for real-time fraud detection, contributing to improved security and trust in modern banking systems.