Omar Mrani-Zentar
PhD researcher in stochastic control and machine learning
I work on the mathematical theory of decision-making under uncertainty, with a focus on decentralized stochastic control, data-driven control, and reinforcement learning. My doctoral research contributes to the theory of learning and optimization in complex environments.
Research Interests
- Stochastic Control Robustness and learning in complex systems.
- Machine Learning Reinforcement learning, data-driven control and learning, adaptive control.
- Decentralized stochastic control Multi-agent problems with a focus on challenges related to partial information and information structures.
- Probability Theory and Applications Kernel topologies, non-linear filter stability in multi-agent systems.
Journal Publications
Published
- Mrani-Zentar, O., & Yüksel, S. (2026). Approximations and learning for decentralized stochastic control and near optimal finite window policies. To appear in IEEE Transactions on Automatic Control. arXiv:2605.00160
- Bluman, G. W., Mrani-Zentar, O., & Finlay, D. (2018). Composition of Lie Group Elements from Basis Lie Algebra Elements. Journal of Nonlinear Mathematical Physics, 25(4), 528–557. doi:10.1080/14029251.2018.1503398
Under revision
- Mrani-Zentar, O., & Yüksel, S. (2026). Kernel Metrics and Learning for Borel MDPs: Identifiability and Adaptive Control. Under second review at SIAM Journal on Control and Optimization. arXiv:2503.11058
- Mrani-Zentar, O., Simpson, R., et al. (2025). Non-sequential decentralized stochastic control revisited: Causality and static reducibility. Under revision at IEEE Transactions on Automatic Control. arXiv:2211.03949
Conference Proceedings
- Mrani-Zentar, O., & Yüksel, S. (2024). Decentralized stochastic control in Borel spaces: Centralized MDP reductions, near optimality of finite window local information, and Q-learning. In Proceedings of the 63rd IEEE Conference on Decision and Control (CDC), pp. 8435–8440, Milan, Italy, December 16–19, 2024. IEEE. IEEE Xplore
Preprints
- Mrani-Zentar, O., & Yüksel, S. (2025). Centralized reduction of decentralized stochastic control models and their weak-Feller regularity. arXiv:2408.13828
Thesis
- Mrani-Zentar, O. (2019). Low-dimensional Lie algebras and control theory (Master's thesis). University of British Columbia. UBC Library
Selected Talks
- Mrani-Zentar, O. (2024, December). Decentralized stochastic control in Borel spaces: Centralized MDP reductions, near optimality of finite window local information, and Q-learning. 63rd IEEE Conference on Decision and Control (CDC), Milan, Italy.
- Mrani-Zentar, O. (2024, December). Learning and Approximations for Multi-Agent Team Problems with Borel Spaces under Decentralized Information. Large Population Teams: Control, Equilibria & Learning, 2024 CDC full-day workshop.
- Mrani-Zentar, O. (2025, May). Existence, Approximations, and Q-learning for Decentralized Stochastic Control with Standard Borel Spaces. 10th Meeting on System and Control Theory (MSCT), HEC Montréal, Montréal, Québec, Canada, May 7–9, 2025.
Teaching
- MATH/MTHE 337: Introduction to Stochastic Operations Research, Queen's University, Winter 2026