OWASP Cork - Adversarial machine learning

**** Meeting link will be added closer to the date and accessible to those who RSVP only ****

Talk 1: Deep malware obfuscation
Speaker: Saeid Rezaei - Machine learning researcher at University College Cork

Deep learning has been widely utilized in many practical applications and achieved remarkable commercial success. It can be used in malware obfuscation engines which are used to secure digital rights management against hackers, and bypass anti-malware programs. In the real world scenarios, there are some constraints that should be considered during malware obfuscation. Using deep learning can help to select the best samples which are satisfying these constraints. In this presentation, we will show how can we use deep learning techniques to improve the performance of obfuscator engines against malware detection techniques.

Talk 2: Adversarial machine learning (AML)
Speakers: Dr. Sorcha Healy Principal Engineer, Lead Data Scientist at McAfee and Dr. Catherine Huang - Principal Engineer and Senior Staff Data Scientist at McAfee

In this talk Sorcha and Catherine will review the general attack landscape against ML and ML security providers. Understanding the fundamentals behind what a machine learning model does and how it works in combination with the Mitre Adversarial Matrix can provide some sensible MLOps practices to protect the components of a ML system. We will review adversarial machine learning (AML) use cases and demonstrate McAfee’s applied research in AML in protecting our customers against this hostile landscape.