COMPARATIVE ANALYSIS OF MACHINE LEARNING ALGORITHMS FOR BANKING FRAUD DETECTION: A STUDY ON PERFORMANCE, PRECISION, AND REAL-TIME APPLICATION

This study investigates the application of machine learning algorithms for fraud detection in the banking sector, addressing the increasing sophistication of fraudulent activities in digital banking. A comparative analysis was conducted on various models, including logistic regression, decision trees, random forests, support vector machines, neural networks, and ensemble methods. Performance metrics such as precision, recall, F1-score, and AUC-ROC were used to evaluate model effectiveness. Results indicate that ensemble models, specifically the stacked ensemble, outperformed other algorithms in balancing precision and recall, thus minimizing false positives and false negatives. These models demonstrated superior accuracy and adaptability to complex fraud patterns, making them particularly suitable for real-time fraud detection. However, challenges related to model interpretability and data quality highlight the need for further research on explainable AI and unsupervised learning approaches. This study underscores the promise of machine learning as a strategic solution for enhancing fraud detection in banking, offering a path to more robust and responsive financial security measures.