CS285

Abstraction: UC Berkeley graduate deep reinforcement learning course lecture index

Key points:

• 23-lecture course covering policy gradients, actor-critic, Q-functions, model-based RL, exploration, offline RL, and RL theory
• Lecture recordings available on YouTube playlist; slides published as PDFs
• Topics span imitation learning, value function methods, inverse RL, meta-learning, and transfer/multi-task learning
• Includes a TensorFlow and neural nets review session (Lecture 3) with accompanying Jupyter notebook
• Covers advanced topics: connection between inference and control, probability/variational inference primer, and open problems

Connections: Uc Berkeley · Reinforcement Learning · Policy Gradients · Model Based RL

Source: http://rail.eecs.berkeley.edu/deeprlcourse-fa20/