Adversarial ML

Topics Covered

• Introduction to Adversarial Machine Learning
• Attacking Malware Detection

References and Reading

The following resources were summarized and discussed in section:

• The paper Android HIV: A Study of Repackaging Malware for Evading Machine-Learning Detection approaches the problem of tweaking Android APKs to evading detection by two recent ML-powered Malware detection algorithms.

• The paper MAMADROID: Detecting Android Malware by Building Markov Chains of Behavioral Models designs a Malware detection algorithm based on sequences of API calls. Besides being the target of the Android HIV paper, this paper is a good example of experimental design in Malware detection to follow in your project proposals.

• A thorough set of references and definitions for the subject of adversarial machine learning can be found in A Taxonomy and Terminology of Adversarial Machine Learning.

Tasks

Continue working on your Hindroid replication project.

Participation

Please complete the ‘default’ participation assignment, answering questions about the reading.