Learn2Evade: Learning-based Generative Model for Evading PDF Malware Classifiers (to appear),

Ho Bae † , Younghan Lee † , Yohan Kim, Uiwon Whang , Sungroh Yoon * , and Yunheung Peak *

∗ : Correspondence should be addressed to S. Yoon and Y.Peak.

† : These authors contributed equally to this work.’

IEEE Transactions on Artificial Intelligence

본 논문은 PDF malware을 방어하는 Anti-Virus engine들에 대하여 높은 공격 성공률을 보여주는 generative model을 활용한 새로운 공격 방법론을 제안하였습니다.