Dynamic Ransomware Detection through Adaptive Anomaly Partitioning Framework
Allan Hatt, Christopher Blackwood, Benjamin Lockhart, Jonathan Ravenscroft, Nicholas Wainwright · 2024
The escalating frequency and sophistication of cyber threats require the development of more advanced detection mechanisms. The Adaptive Anomaly Partitioning Framework introduces a novel approach to ransomware detection through its dynamic partitioning mechanism, which enables the system to adapt to emerging threats without human intervention. This framework employs a modular design that facilitates continuous learning, thereby enhancing its efficacy in identifying and mitigating ransomware activities. Empirical evaluations demonstrate high detection accuracy and precision across diverse datasets, with scalability assessments indicating its suitability for large-scale network environments. The framework's adaptability to various ransomware variants and its low false positive rate demonstrate its robustness and reliability. These attributes position the Adaptive Anomaly Partitioning Framework as a significant advancement in cybersecurity defenses, offering a promising solution for mitigating the impact of ransomware attacks on critical systems and data.