Vasilis holds a Meng in Electrical and Computer Engineering from the National Technical University of Athens with a specialization in Computer Science. He has research and work experience in the development and analysis of machine learning algorithms. In 2020 he was an intern at NCSR “Demokritos”, where he applied deep learning techniques to satellite imagery problems. From September 2022 to August 2023 he worked as a research assistant at the Max Planck Institute for Software Systems, focusing on reinforcement learning and robustness problems.

RESEARCH INTERESTS

Vasilis' research interests include optimization and statistical learning theory. Specifically, his research will focus on areas such as learning in games and computation of equilibria.

PhD research on "Learning in Non-Convex Games: Algorithms, Dynamics, and Convergence Guarantees"
Abstract: The main goal of this PhD will be to design implementable algorithms capable of negotiating the fundamental obstructions faced by gradient methods in min-max problems (or, more generally, equilibrium problems) without a convex structure. This will involve combining techniques from zeroth-order optimization and multi-armed bandits (stochastic or non-stochastic) with the analysis of standard gradient methods for regularized learning in games.