Improving robustness of ML classifiers against realizable evasion attacks using conserved features

Keyphrases

• Machine Learning Classifiers 100%
• Evasion Attack 100%
• Feature Space Model 100%
• Malware 66%
• Malware Detection 66%
• Robust Machine Learning 66%
• Insecurity 33%
• Performance Improvement 33%
• Machine Learning 33%
• Attacker 33%
• Security Applications 33%
• Adversary 33%
• Content-based 33%
• Machine Learning Techniques 33%
• Machine Learning Models 33%
• Intrusion Detection 33%
• Learned Features 33%
• Machine Learning Robustness 33%
• PDF Malware 33%

Computer Science

• Machine Learning 100%
• Malware 50%
• Feature Space 50%
• Attackers 16%
• Security Application 16%
• Machine Learning Technique 16%
• Malware Detection 16%
• Intrusion Detection 16%