TODO

1. 정상, 악성 파일 비율 맞춰서 비교
2. URL 배제
3. Exploit --> OBJ Length, OBJ Entropy, (관련 논문)
4. Data Augmentation 방법 모색

Papers

• Davide Maiorca, Battista Bigg,Giogio Giacinto, “Towards Adversarial Malware Detection: Lessons Learned from PDF-based Attacks”
• CADE
• Nedim Šrndic, Pavel Laskov, “Detection of Malicious PDF Files Based on Hierarchical Document Structure”
• Nedim Šrndic, Pavel Laskov, “Hidost: a static machine-learning-based detector of malicious files”, EURASIP Journal on Information Security volume, 2016
• On Training Robust PDF Malware Classifiers