معرفی
Anas Barakat is a Research Fellow at the Singapore University of Technology and Design, where he develops theory and algorithms for learning in strategic, structured, and dynamic environments. His work integrates reinforcement learning, game theory, online learning, stochastic optimization, and dynamical systems to build robust multi-agent learning systems.
- PhD: Applied Mathematics and Computer Science, 2021, Institut Polytechnique de Paris (Télécom Paris)
- MSc: Data Science, 2018, Université Paris Saclay
- MSc: Applied Mathematics and Computer Science, 2018, Télécom Paris
His research focuses on multi-agent learning in strategic environments, behaviorally aligned reinforcement learning (e.g., incorporating psychological biases), and adaptive optimization via dynamical systems analysis. He explores feedback loops, non-stationary objectives, and structured games like Markov potential games and zero-sum linear quadratic games.
Recent publications (2025–2023) span multi-agent control, symmetric cone games, policy gradient frameworks, and Adam algorithm convergence. His work appears in venues like IEEE CDC, NeurIPS, ICML, and AISTATS, with preprints on arXiv.
At ETH Zurich (2022–2024), he taught Optimization for Data Science and Foundations of Reinforcement Learning, and at Télécom Paris (2018–2021), he assisted courses in machine learning and optimization.
