Omar Saadi

Research Interests 

• Zeroth-order optimization methods

• Markov decision processes, Zero-sum stochastic games, reinforcement learning

• Dynamic programming, operations research, optimal control

• Accelerated algorithms for fixed point problems

• Non-linear Perron-Frobenius theory, tropical linear regression, tropical low-rank approximation

Publications and Preprints

• E. M. Chayti, T. E. B. El Kadi, O. Saadi and M. Jaggi. Stochastic Optimization with Random Search. 

• T. E. B. EL KADI and O. Saadi. On the Almost Sure Convergence of the Stochastic Three Points Algorithm. In The Thirteenth International Conference on Learning Representations (ICLR 2025).

• M. Akian, S. Gaubert, Y. Qi and O. Saadi. Tropical linear regression and mean payoff games: or, how to measure the distance to equilibria, SIAM Journal on Discrete Mathematics, 2023, vol. 37, no 2, p. 632-674. 

• M. Akian, S. Gaubert, Z. Qu and O. Saadi. Multiply Accelerated Value Iteration for Non-Symmetric Affine Fixed Point Problems and application to Markov Decision Processes, SIAM Journal on Matrix Analysis and Applications, 2022, vol. 43, no 1, p. 199-232.   

• M. Akian, S. Gaubert, Z. Qu and O. Saadi. Solving ergodic Markov decision processes and perfect information Zero-sum stochastic games by variance reduced deflated value iteration, in Proc. of the 58th Conference on Decision and Control, 2019. 

• M. Akian, S. Gaubert, Y. Qi and O. Saadi. A geometric approach of best tropical low rank approximation. Working paper. 

Conferences and talks

• METIS’2025 Spring School: Trends in Machine Learning, Rabat, Morocco, May 2025. 

• The Thirteenth International Conference on Learning Representations (ICLR 2025), Singapore, April 2025. 

• Math&Maroc Competition event for top first year university students, Benguerir, Morocco, April 2023. 

• AI for Science Workshop, Rabat, Morocco, December 2022. 

• 22ème Congrès de la société Française de Recherche Opérationnelle et d'Aide à la Décision (ROADEF2021), Mulhouse, France, April 2021.

• French Days on Optimization and Decision Science, SMAI-MODE, Palaiseau, France, September 2020.

• Journée des Doctorants 2A, CMAP, Ecole Polytechnique, Palaiseau, France, June 2020.

• 58th IEEE Conference on Decision and Control, Nice, France, December 2019.

• PGMO Days, EDF Lab Paris-Saclay, Palaiseau, France, December 2019.

• French-German-Swiss Conference on Optimization, Nice, France, Septembre 2019.

• Workshop - Theory and Algorithms in Graph and Stochastic Games, Mons, Belgium, March 2019.

Teaching 

• UM6P, Moroccan Center for Game Theory, PhD Program: Operational Research (Fall 2025).

• UM6P, Computer Science School, Engineering cycle (2nd year): Operational Research (Fall 2025).

• UM6P, Computer Science School, Engineering cycle (1st year): Graph Algorithms (Fall 2025).

• UM6P, Moroccan Center for Game Theory, PhD Program: Operational Research (Fall 2024).

• UM6P, Computer Science School, Engineering cycle (2nd year): Operational Research (Fall 2024).

• UM6P, Computer Science School, Engineering cycle (1st year): Graph Algorithms (Fall 2024).

• UM6P, Computer Science School, Engineering cycle (1st year): Graph Algorithms (Fall 2023).

• UM6P, QFM Master (M1): M113 - Optimization & Operations Research (Fall 2023).

• UM6P, LSD Bachelor (L2): M24 - Operations Research (Spring 2023).

• UM6P, QFM Master (M1): M113 - Optimization & Operations Research (Fall 2022). 

• Ecole Polytechnique (third-year students, 3A): MAP557 - Operations research: mathematical aspects and applications (Fall 2018, 2019 and 2020).

• Ecole Polytechnique (second-year students, 2A): MAP473D - Modal : Simulation numérique aléatoire (SNA) autour des événements rares (Spring 2019, 2020 and 2021).