MACHINE LEARNING FOR FRAUD DETECTION IN DIGITAL BANKING: A SYSTEMATIC LITERATURE REVIEW

Abstract

This systematic literature review examines the role of machine learning in fraud detection within digital banking, synthesizing evidence from 118 peer-reviewed studies and institutional reports. Following the PRISMA guidelines, the review applied a structured identification, screening, eligibility, and inclusion process to ensure methodological rigor and transparency. The findings reveal that supervised learning methods, such as decision trees, logistic regression, and support vector machines, remain the dominant paradigm due to their interpretability and established performance, while unsupervised anomaly detection approaches are increasingly adopted to address novel fraud patterns in highly imbalanced datasets. Deep learning architectures, particularly recurrent and convolutional neural networks, have emerged as transformative tools capable of modeling sequential transaction data and detecting complex fraud typologies, though challenges of interpretability and real-time deployment persist. Hybrid models that combine supervised, unsupervised, and deep learning strategies demonstrate superior adaptability and detection accuracy, highlighting their potential as convergent solutions. The review further underscores the importance of evaluation metrics—precision, recall, F1-score, and PR-AUC—as well as explainability frameworks like SHAP and LIME, which ensure that models are both statistically robust and operationally transparent. Cross-regional analysis shows that regulatory environments and institutional capacities shape methodological adoption: the European Union emphasizes compliance under PSD2 and GDPR, North America leverages fintech partnerships and data-driven innovation, and emerging economies rely heavily on infrastructure and governance maturity. Despite methodological advances, gaps remain in reproducibility, robustness under distributional shift, and theoretical integration with criminological and governance frameworks.