BYSD/rl_at_ammi

Reinforcement Learning Tutorials for the African Masters of Machine Intelligence

Reinforcement Learning at AMMI

This repo includes lecture slides, notebooks and other material for the RL week at AMMI, the African Masters of Machine Intelligence

Policy/Value Iteration

• Day 1 Slides
• Policy/Value Iteration exercises
• Policy/Value Iteration solutions

Extra:

• Supplementary: Policy/Value Iteration in Matrix form
• RL book Chapter 4, Dynamic Programming
• Dynamic Programming lecture video

Monte Carlo and TD learning

• Day 3 Slides
• Monte Carlo / TD exercises
• Monte Carlo / TD solutions

Function Approximation

• Day 5 Slides

Policy Gradient - REINFORCE

• Day 6 Slides
• REINFORCE exercise
• REINFORCE solution
• REINFORCE with learned Baseline exercise
• REINFORCE with learned Baseline solution
• Actor-Critic
• Actor-Critic solution

Extra:

• Policy Gradient Algorithms
• Policy Gradient Explained - Blog Post
• Policy Gradient lecture video

Models and Hierarchy

• Day 8 Slides

Python and Numpy

• Python Basics

Extra:

• Python crash course
• Numpy crash course
• How to build your own Neural Network from scratch in Python

External Resources

• Reinforcement Learning: An Introduction, by Sutton and Barto, 2018. This is the canonical RL book which has everything you need to learn RL.
• RL Course Lectures by David Silver, videos, slides