2 Veni Grants for the ELLIS Unit
Signal Processing and Learning from Higher-Order Network Dynamics
Dr. Elvin Isufi, Electrical Engineering, Mathematics & Computer Science (EEMCS)
Networks, such as those that distribute water to our homes or our brain, generate streams of data according to their topology but conventional processing techniques do not fully capture their complex dependencies. The research will investigate novel techniques to better leverage the network structure for processing these data so as to detect more accurately brain anomalies, forecast future water demands, and make them deployable to a large-scale setting.
Adaptive Algorithms for Non-Stationary Reinforcement Learning
Dr. Julia Olkhovskaia, Electrical Engineering, Mathematics & Computer Science (EEMCS)
Reinforcement Learning (RL) is revolutionizing automation tasks such as autonomous driving and managing smart power grids. RL stands out for its unique ability to actively learn from interactions and adapt to new data. However, in rapidly changing environments, it is struggling to perform in large-scale problem settings. My research aims to overcome this by developing advanced RL algorithms that are not only adaptive but can also process a vast amount of environmental data in real time. This approach is bridging the gap between theoretical RL models and their practical, real-world applications.
