MEME Algorithm's Impact on Adversarial Malware Generation and Model Evasion
2024-4-19 01:0:3 Author: hackernoon.com(查看原文) 阅读量:0 收藏

Too Long; Didn't Read

MEME's experiments evaluated its performance against top malware detection systems like Ember, Sorel-LGB, Sorel-FFNN, and Microsoft Defender. The comparison included state-of-the-art techniques like PPO, MAB, and GAMMA, focusing on adversarial malware generation. Constraints on modifications, running time, query budgets, and evaluation metrics standardized the experiments for fair comparison and robust assessment of MEME's efficacy in model evasion.


文章来源: https://hackernoon.com/meme-algorithms-impact-on-adversarial-malware-generation-and-model-evasion?source=rss
如有侵权请联系:admin#unsafe.sh