A ransomware-risk classifier trained on real victim records plus synthetic safe and unsafe samples reaches 99% test accuracy, but the evaluation is circular because the synthetic labels are built from the same features the model uses.
Cyber threat intelligence sharing: Survey and research directions,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.CR 1years
2025 1verdicts
REJECT 1roles
background 1polarities
support 1representative citing papers
citing papers explorer
-
Assessing and Prioritizing Ransomware Risk Based on Historical Victim Data
A ransomware-risk classifier trained on real victim records plus synthetic safe and unsafe samples reaches 99% test accuracy, but the evaluation is circular because the synthetic labels are built from the same features the model uses.