Introduction to learning-based control methodologies. Review of fundamental regression concepts. Sparse Identification of Nonlinear Dynamics (SINDy). Real-time system identification. Classical adaptive control principles, including both indirect methods and direct architectures (e.g. MRAC). Parameter convergence analysis. Advanced state estimation techniques based on stochastic approaches, including Variational Inference. Learning control policies from demonstration data (Imitation Learnin. In the second part, the course focuses on stochastic optimal control and Reinforcement Learning (RL). Dynamic programming is studied for both finite- and infinite-horizon problems, followed by the transition to model-free reinforcement learning methods. Actor–critic algorithms for continuous control (e.g., DDPG) and stochastic policy-gradient methods (e.g., PPO, TRPO) are presented. A significant component of the course consists of computational exercises and a final project.