Expected SARSA update rule

In this exercise, you'll implement the Expected SARSA update rule, a temporal difference model-free RL algorithm. Expected SARSA estimates the expected value of the current policy by averaging over all possible actions, providing a more stable update target compared to SARSA. The formulas used in Expected SARSA can be found below.

Image showing the mathematical formula of the expected SARSA update rule.

The numpy library has been imported as np.

This exercise is part of the course

Reinforcement Learning with Gymnasium in Python

Exercise instructions

Calculate the expected Q-value for the next_state.
Update the Q-value for the current state and action using the Expected SARSA formula.
Update the Q-table Q supposing that an agent takes action 1 in state 2 and moves to state 3, receiving a reward of 5.

Hands-on interactive exercise

Have a go at this exercise by completing this sample code.

def update_q_table(state, action, next_state, reward):
  	# Calculate the expected Q-value for the next state
    expected_q = ____
    # Update the Q-value for the current state and action
    Q[state, action] = ____
    
Q = np.random.rand(5, 2)
print("Old Q:\n", Q)
alpha = 0.1
gamma = 0.99

# Update the Q-table
update_q_table(____, ____, ____, ____)
print("Updated Q:\n", Q)

Edit and Run Code

This exercise is part of the course

Reinforcement Learning with Gymnasium in Python

AdvancedSkill Level

4.8+

Start Course for Free

Dive into the exciting world of Reinforcement Learning (RL) by exploring its foundational concepts, roles, and applications. Navigate through the RL framework, uncovering the agent-environment interaction. You'll also learn how to use the Gymnasium library to create environments, visualize states, and perform actions, thus gaining a practical foundation in RL concepts and applications.

Exercise 1: Fundamentals of reinforcement learning Exercise 2: What is Reinforcement Learning?Exercise 3: RL vs. other ML sub-domains Exercise 4: Scenarios for applying RL Exercise 5: Navigating the RL framework Exercise 6: RL interaction loop Exercise 7: Episodic and continuous RL tasks Exercise 8: Calculating discounted returns for agent strategies Exercise 9: Interacting with Gymnasium environments Exercise 10: Setting up a Mountain Car environment Exercise 11: Visualizing the Mountain Car Environment Exercise 12: Interacting with the Frozen Lake environment

Delve deeper into the world of RL focusing on model-based learning. Unravel the complexities of Markov Decision Processes (MDPs), understanding their essential components. Enhance your skill set by learning about policies and value functions. Gain expertise in policy optimization with policy iteration and value Iteration techniques.

Exercise 1: Markov Decision Processes Exercise 2: Custom Frozen Lake MDP components Exercise 3: Exploring state and action spaces Exercise 4: Transition probabilities and rewards Exercise 5: Policies and state-value functions Exercise 6: Defining a deterministic policy Exercise 7: Computing state-values for a policy Exercise 8: Comparing policies Exercise 9: Action-value functions Exercise 10: Computing Q-values Exercise 11: Improving a policy Exercise 12: Policy iteration and value iteration Exercise 13: Applying policy iteration for optimal policy Exercise 14: Implementing value iteration

Embark on a journey through the dynamic realm of Model-Free Learning in RL. Get introduced to to the foundational Monte Carlo methods, and apply first-visit and every-visit Monte Carlo prediction algorithms. Transition into the world of Temporal Difference Learning, exploring the SARSA algorithm. Finally, dive into the depths of Q-Learning, and analyze its convergence in challenging environments.

Exercise 1: Monte Carlo methods Exercise 2: Episode generation for Monte Carlo methods Exercise 3: Implementing first-visit Monte Carlo Exercise 4: Implementing every-visit Monte Carlo Exercise 5: Temporal difference learning Exercise 6: Implementing the SARSA update rule Exercise 7: Solving 8x8 Frozen Lake with SARSA Exercise 8: Q-learning Exercise 9: Implementing Q-learning update rule Exercise 10: Solving 8x8 Frozen Lake with Q-learning Exercise 11: Evaluating policy on a slippery Frozen Lake

Dive into advanced strategies in Model-Free RL, focusing on enhancing decision-making algorithms. Learn about Expected SARSA for more accurate policy updates and Double Q-learning to mitigate overestimation bias. Explore the Exploration-Exploitation Tradeoff, mastering epsilon-greedy and epsilon-decay strategies for optimal action selection. Tackle the Multi-Armed Bandit Problem, applying strategies to solve decision-making challenges under uncertainty.

Exercise 1: Expected SARSA Exercise 2: Expected SARSA update rule

Current Exercise

Exercise 3: Applying Expected SARSA Exercise 4: Double Q-learning Exercise 5: Implementing double Q-learning update rule Exercise 6: Applying double Q-learning Exercise 7: Balancing exploration and exploitation Exercise 8: Defining epsilon-greedy function Exercise 9: Solving CliffWalking with epsilon greedy strategy Exercise 10: Solving CliffWalking with decayed epsilon-greedy strategy Exercise 11: Multi-armed bandits Exercise 12: Creating a multi-armed bandit Exercise 13: Solving a multi-armed bandit Exercise 14: Assessing convergence in a multi-armed bandit Exercise 15: Congratulations!