Enhancing Fraud Detection in Payment Systems Using Explainable AI and Deep Learning Techniques

The evolving nature of financial fraud necessitates advanced detection methods beyond traditional rule-based systems and statistical approaches. This study developed a deep learning-based model for financial fraud detection, incorporating explainable AI (XAI) techniques to enhance interpretability. The model achieved a precision of 99.995% in identifying financial fraud, demonstrating significant improvement over traditional methods. By integrating XAI, using SHAP was further effective in identifying influential features in the model. Our findings indicate that the model not only enhances fraud detection accuracy but also offers stakeholders a clear understanding of fraud identification factors, potentially leading to more robust and trustworthy financial systems. However, this study highlights the need for further research on enhancing the interpretability of deep learning models in fraud detection, emphasizing the importance of balancing the model complexity with explainability to have a resilient financial security system.