Lecture 12, 2025; Training of cost functions, approximation in policy space, policy gradient methods

This site also contains complete PDF of related textbooks by Bertsekas:
"A Course in Reinforcement Learning", 2nd edition, 2025
"Lessons from AlphaZero for Optimal, Model Predictive, and Adaptive Control," 2022
"Abstract Dynamic Programming", 3rd edition, 2022
"Rollout, Policy Iteration, and Distributed Reinforcement Learning," 2020
Lecture given by Dimitri Bertsekas
Training of cost functions and policies, incremental gradient methods, approximation in policy space, policy gradient methods, random search. Pitfalls of off-line training. Final remarks and an overview of the course,

[Imanol]

Professor Bertsekas, thank you for making these lectures and notes available. As an undergraduate student passionate about optimization and reinforcement learning, I greatly admire your work and hope to learn from you in the future