Ransomware Detection Using Algorithmic Entropy-Based Neural Correlation Analysis
Charles Burck, Malachi Frantzen, Benjamin Harris, Mathias Sorokina, Richard Wright · 2024
The proliferation of malicious encryption-based cyberattacks has posed significant challenges to existing detection systems, which often struggle to keep pace with rapidly evolving threat landscapes. Introducing a computational framework that combines algorithmic entropy with neural correlation analysis addresses critical gaps in detection capabilities by enabling the identification of complex and previously untraceable malicious patterns. The proposed approach achieves high accuracy and efficiency through a synergy of entropy-based evaluations and advanced neural modeling, offering an automated and scalable solution to contemporary challenges. Comprehensive experiments demonstrate the framework's superior detection rates and minimal false positives across diverse datasets, highlighting its adaptability to varying operational environments and its potential to redefine cybersecurity methodologies. These results establish a robust foundation for integrating advanced computational paradigms into real-time threat detection systems, significantly enhancing their resilience against dynamic and sophisticated ransomware tactics.