Explain the key elements of reinforcement learning: policies, reward signals, value functions, and models.
Formulate problems as Markov decision processes and identify states, actions, and rewards.
Apply dynamic programming methods to compute optimal policies in known environments.
Implement and analyze multi-arm bandit algorithms, including epsilon-greedy and UCB.
Use Monte Carlo methods for policy evaluation and control without a model.
Balance exploration and exploitation in reinforcement learning tasks.
