BriefGPT.xyz
Sep, 2019
对恶意软件分类中敌对样本及其防御策略的有效性研究
Effectiveness of Adversarial Examples and Defenses for Malware Classification
HTML
PDF
Robert Podschwadt, Hassan Takabi
TL;DR
本论文研究了在恶意软件检测任务中制造对抗样本和对抗样本对抗技术的不同方法，并比较了它们对多个数据集的有效性。
Abstract
artificial neural networks
have been successfully used for many different
classification tasks
including
malware detection
and distinguish
→