An Empirical Evaluation of Supervised Learning Methods for Network Malware Identification Based on Feature Selection

Engineering & Materials Science

• Malware 100%
• Supervised learning 96%
• Random forests 96%
• Feature extraction 70%
• Logistic regression 67%
• Machine learning 44%
• Packet networks 35%
• Multilayer neural networks 31%
• Decision trees 29%
• Principal component analysis 28%
• Statistical methods 25%
• Learning algorithms 25%
• Support vector machines 24%
• Classifiers 21%
• Neural networks 18%
• Experiments 10%