Jonas Wagner machine_learning_and_malware.pdf

Machine learning is the science of getting computer to learn and act like humans do, and improve their learning over time in autonomous fashion, by feeding them data and informatin in the form of observations and real-world interactions.

• Autonomy and scale easily, are better than humans at specific learned tasks.
• Adaptability and flexebility
  • Attackers change their behavior, so should defenders
  • Autonomous reajustment based on new data
• Make certain techinques possible that would not be feasible without machine learning.

• High data dependency
  • The system is only as good as the data it was trained on
  • Easy to overfit
• Good models are not necessarily good for production
  • Models can be too slow and complex
• Blackbox results are difficult to interpret

• learn a function that maps input to output
• The inference process requires labeled data
• We are giving features and labels
• used for classification and regression