Application of Advanced Machine Learning Methods in Financial Fraud Detection

The identification and prevention of financial fraud represent one of the most critical challenges facing organizations and financial institutions, with significant implications for the health and stability of economic systems. The present study investigates the application of advanced machine learning methods in detecting financial fraud, aiming to evaluate and compare the performance of several machine learning models in this domain. In this research, decision tree, artificial neural networks, support vector machines, gradient boosting, and random forest methods were applied to financial data from companies listed on the Tehran Stock Exchange. After conducting the necessary preprocessing steps, the models were trained. The results indicate that the gradient boosting model, with an accuracy of 95%, a sensitivity of 92%, and an F1-score of 0.94, delivered the best performance in detecting financial fraud. Additionally, artificial neural networks and decision tree models demonstrated acceptable performance with accuracies of 90% and 80%, respectively. Unlike previous studies that typically focused on a limited set of algorithms, this study adopts a comprehensive and multi-model approach, enabling a precise and practical comparison of various algorithms’ performance. The use of real financial data and rigorous preprocessing enhanced the validity and generalizability of the findings. The results of this study suggest that employing advanced machine learning methods—particularly the gradient boosting algorithm—can significantly enhance the accuracy and sensitivity of financial fraud detection compared to traditional methods, and can serve as an effective and practical solution in fraud detection systems. This multi-model and application-oriented approach highlights the study’s main innovation in improving the precision and efficiency of financial fraud detection using state-of-the-art machine learning technologies, and constitutes a meaningful step toward strengthening security and trust in financial systems.