fa25-cs540 cs540
Lecture 24: Reinforcement Learning II Lecture 24: Reinforcement Learning II
Content locked
The catalog is public. Reading this content requires approved access.
Sign in to readLecture 24: Reinforcement Learning II
Introduction
Review of Reinforcement Learning
The Agent-World Model
Markov Decision Process (MDP)
Value Function
Discounting Rewards
Values and Policies
Obtaining the Optimal Policy
Bellman Equations
Example: Computing Values
Value Iteration
Q-Learning
The Q-Function
Q-Learning Update Rule
Q-Learning Algorithm
Example: Q-Learning Step-by-Step
Exploration vs. Exploitation
Exploration
Exploitation
Epsilon-Greedy Policy
SARSA: An Alternative Update Rule
Q-Learning Details
Terminal States and Episodes
Deep Q-Learning
Architecture
Training
Practice Problems
Key Takeaways
Suggested Reading