Threat Modeling Research and Machine Learning
Current software engineering research is investigating some connections between security analysis and machine learning – both the potential vulnerability of machine learning applications as well as using machine learning in threat modeling analysis.
This talk will focus on recent threat modeling research as it relates to machine learning. After briefly revisiting our prior threat modeling research, new results from a 2018 student project on machine learning will be discussed. In this project, students assessed the robustness of machine learning models against adversarial examples. Recently, we have been considering the use of machine learning to identify attacker types in specific domains. So, on the one hand, we examined whether machine learning models are vulnerable to attack, and on the other hand, whether machine learning can help to identify attacker types.
Date and Time
Location
Hosts
Registration
- Date: 18 Mar 2021
- Time: 08:00 PM to 09:30 PM
- All times are (GMT-05:00) US/Eastern
-
Add Event to Calendar
Speakers
Nancy Mead of Carnegie Mellon University
Biography:
Nancy Mead is a Fellow of the Software Engineering Institute (SEI), and an Adjunct Professor of Software Engineering at Carnegie Mellon University. Her research areas are security requirements engineering and software assurance curricula. Prior to joining the SEI, Nancy was a senior technical staff member at IBM Federal Systems, where she spent most of her career in the development and management of large real-time systems. Nancy has more than 150 publications and invited presentations. She is a Life Fellow of the IEEE, a Distinguished Member of the ACM, and was named the 2015 Distinguished Educator by IEEE TCSE. She received her PhD in mathematics from the Polytechnic Institute of New York.