A Crash Course in Reinforcement Learning
Outline
Learning To Act
RL at a glance
Acting in Action
Markov Decision Processes
Model Parameters: transition probabilities
Returns and discounting
RL Concepts
Policies and Values
Optimal Policies
The action value function
Model-Based Learning
Model-free Learning
Model-free Learning: What are the data?
Temporal Difference Learning
Model-Free Learning Example

Model-free Learning

Directly learn optimal policy π* (policy iteration) Directly learn optimal value function V*. Directly learn optimal action-value function Q*. All of these functions can be implemented in a neural network. NN learning = reinforcement learning