RANDS: A Machine Learning-Based Anti-Ransomware Tool for Windows Platforms
Hiba Zuhair, Selamat Ali · Frontiers in artificial intelligence and applications · 2019
Zero-day ransomware still threaten users' and enterprises' survival in the cyber-space by disturbing electronic amenities, damaging information systems, and causing data and money losses. Established anti-ransomware techniques are trying to mitigate this security issue, however they are lacking to identify ransomware families effectively without real-time performance overhead. Thus, this paper provides a multi-tier anti-ransomware tool (RANDS) performs via windows platform through three tiers: ransomware analysis tier, learning tier and detection tier. RANDS hybridizes the decisive functions of two machine learning algorithms (Naïve Bays and Decision Tree) to holistically analyze ransomware traits, and accurately classify ransomware families. The prototype implementation of RANDS shows its classification capability against ransomwares with (96.27%) as average accuracy rate and (1.32%) of average mistake rate throughout real-time assessment.