I work in the field of self-aware computing systems, especially in the domains of embedded and high-performance computing. Self-awareness enables a system to continuously observe and analyze its own state and the external environment to adapt accordingly and increase the system's utility. Typically, such a behavior is required if the system will face situations during operation which cannot be foreseen at design time. Hence, self-aware systems are often equipped with on-line machine learning techniques to learn on their own what actions are called for. I place my focus on the application of Learning Classifier Systems (LCS), which combine reinforcement learning with genetic algorithms to evolve a human-interpretable rule base.
If some of this sounds interesting and you think about getting involved, take a look at my list of open thesis topics below. In case there is no suitable topic announced and/or you have an own idea, feel free to contact me so we can discuss!