arXiv 2018

Learning to Evade Static PE Machine Learning Malware Models via Reinforcement Learning

arXiv 2018 endgameinc/gym-malware

We show in experiments that our method can attack a gradient-boosted machine learning model with evasion rates that are substantial and appear to be strongly dependent on the dataset.

MALWARE DETECTION