Our research group works on the application of machine learning in the computer security. In particular, we develop methods in the area of system security and application security, as for instance, attack detection or vulnerability discovery in software and embedded devices. Also, the robustness, security, and interpretability of machine learning methods are central to our research.
Explaining and understanding machine learning models is crucial for the reliable use of learning-based systems. In our research, we investigate the potentials and limits of XAI.
A thorough analysis of malware is essential for successfully protecting computer systems. In our research, we provide insights into novel threats and develop assitive tools using machine learning.
We research approaches to increase the robustness and the security of machine learning algrithms and AI. We analyze, harden, and refine ML models.
Software vulnerabilities pose an inherent threat to the security of IT-systems. We develop learning-based methods to find such exploitable flaws before an adversary does.
In our research, we also address the current urgent need of more advanced methods for detecting attacks. Machine learning and self-learning systems are particular promising tool to do so.
To better understand the adversary and his/her capabilities, we also invest in researching novel attack vectors in different fields and applications.