Emmarius Delar
PhD Student in Reinforcement Learningfirstname [.] lastname [at] dauphine.eu
I am a first-year PhD researcher at LAMSADE (Université Paris Dauphine–PSL), advised by Prof. Tristan Cazenave and affiliated with the PR[AI]RIE Institute. My research is fully funded by an ANR project led by Orange.
About Me
My research lies at the intersection of Reinforcement Learning, Monte Carlo methods, and Combinatorial Optimisation on graphs. Concretely, I develop MCTS-based algorithms to tackle complex graph-structured problems such as Virtual Network Embedding, Boolean models for Gene Regulatory Networks, Graph reconstruction, and Intelligent mesh generation. I am particularly interested in multi-objective RL-based algorithms, and lazy-evaluation strategies that reduce simulation cost without sacrificing performance.
In 2025, I graduated with a Master's degree in Applied Mathematics and Computer Science at PSL Research University. Prior to that, I completed a Bachelor's degree in Applied Mathematics and Computer Science at the University of Rennes.
Outside academia, I am a huge fan of sport — football above all. I follow several leagues but Real Madrid will always be my club, through thick and thin. Growing up as the youngest in a large family, I inherited a wide range of interests from my siblings: cooking, gardening, electronics, tinkering, music, and the occasional video game rabbit hole.
I am always happy to connect with fellow researchers — whether your work overlaps with mine or you are simply navigating the early days of a PhD. Feel free to reach out by email or LinkedIn.
News
Publications
-
[Sept. 2024] Méthodes de réduction de dimension par apprentissage pour l'émulation statistique d'un modèle de dispersion atmosphérique. [report]
- More coming soon!
- Coming soon!