Catalogue · Faculty of Machine Learning
ML 320
Reinforcement Learning
David Silver’s UCL lectures, recorded shortly before his team’s AlphaGo made the subject famous. Markov decision processes, planning, model-free prediction and control, and function approximation — the classical RL canon from the person best placed to teach it. Pair with Sutton & Barto’s free textbook for the full experience.
Syllabus
- Markov decision processes
- dynamic programming
- temporal-difference learning
- policy gradients
- function approximation
Held at UCL & Google DeepMind. The materials remain theirs; the structure is ours. Finished it? Record it in your transcript.