I am an Assistant Professor in the Department of Computer Science at Aarhus University, and a member of the Computational Complexity and Game Theory group. Before that, I was a postdoctoral researcher at École Polytechnique Fédérale de Lausanne and Singapore University of Technology and Design. I received my PhD from National Technical University of Athens in algorithmic game theory.

Research

My research lies at the intersection of game theory, machine learning, and optimization. A big part of my research couples ideas of the above fields to understand dynamics in crucial settings such as markets, blockchain systems, and AI training (see, for example, our work on Ethereum transaction fees). More recently, I have been developing efficient optimization methods for machine learning and exploring the application of deep learning to combinatorial optimization problems.

Awards

• Jan 2025: I have received the Villum Young Investigator Award (9 million DKK / 1.2 million euros) for the project “Developing an online learning approach to multi-agent systems: Game dynamics and equilibria”.

Selected Publications

Game, Market and Price Dynamics

• Semi Bandit dynamics in Congestion Games: Convergence to Nash Equilibrium and No-Regret Guarantees Ioannis Panageas, Stratis Skoulakis, Luca Viano, Xiao Wang and Volkan Cevher, ICML 2023 (oral award).

• $^\clubsuit$Beyond Time-Average Convergence: Near-Optimal Uncoupled Online Learning via Clairvoyant Multiplicative Weights Update: G. Piliouras and R. Simm and S. Skoulakis, **NeurIPS 2022.

• Dynamical analysis of the EIP-1559 Ethereum fee market.: S. Leonardos$^\star$, B. Monnot$^\star$, D. Reijsbergen$^\star$, S. Skoulakis$^\star$ and Georgios Piliouras, AFT 2021.

Equilibrium Computation

• STay-ON-the-Ridge: Guaranteed Convergence to Local Minimax Equilibrium in Nonconvex-Nonconcave Games: Costis Daskalakis, Noah Golowich, Stratis Skoulakis and Manolis Zampetakis, COLT 2023

• The complexity of constrained min-max optimization. Constantinos Daskalakis, Stratis Skoulakis and Manolis Zampetakis, STOC 2021.

Optimization and Machine Learning

• Efficient Efficient Continual Finite-Sum Minimization: I. Mavrothalassitis$^\star$, S. Skoulakis$^\star$, L. Dadi and V. Cevher, ICLR 2024.

• Maximum independent set: Self-training through dynamic programming: L. Brusca$^\star$, L. Quaedvlieg$^\star$, S. Skoulakis$^\star$, G. Chrysos and V. Cevher, NeurIPS 2023.

• Learning to Remove Cuts in Integer Linear Programming: P. Puigdemont$^\star$, S. Skoulakis$^\star$, G. Chrysos and V. Cevher, ICML 2024.