Forecasting Cyber Attacks in Banking and FinTech Platforms: A Review of Data-Driven Approaches

Abstract

It is essential to predict cyber-attacks in financial systems and FinTech platforms to protect financial infrastructures with an ever-advancing threat level. The review will discuss evidence-based solutions that improve the predictive abilities of the Security Information and Event Management (SIEM). Through systematic study of machine learning, deep learning, and statistical models on historical security event data, it identifies key methods for predicting attack vectors such as phishing, DDoS, fraud, and malware infiltration. The exploratory paper identifies feature engineering, anomaly detection, and real-time analytics as components that enhance the accuracy of forecasting. Issues such as imbalanced data, concept drift and latency in large-scale environments are addressed, and new methods such as ensemble learning and adaptive models are mentioned. They also check how the feeds on threat intelligence and the behavioral analytics are integrated into the SIEM systems to prevent risks. The evaluation wraps up with the future research directions, which are scalable, explainable and resilient prediction mechanisms needed to protect banking and FinTech ecosystems.