Reinforcement Learning Exercises

Solutions to RL exercises proposed in the git repo by Denny Britz.

Dynamic Programming

• Policy Evaluation
• Policy Iteration
• Value Iteration

Monte Carlo methods

• Monte Carlo Prediction
• On-policy Monte Carlo Control with e-greedy policy
• Off-policy Monte Carlo Control with Weighted Importance Sampling

Temporal Difference Learning

• SARSA (On-policy TD control)
• Q-learning (Off-policy TD control)

Function approximation

• Q-learning with linear function approximation

Deep Q Learning

• Deep Q learning
• Double Deep Q learning

Policy Gradient

• REINFORCE with baseline
• Actor-Critic with baseline
• Actor-Critic with baseline for continuous action space
• Asynchornous Advantage Actor-Critic (A3C)