We finally kicked off our MLT Reinforcement Learning sessions with an “Intro to RL” study group at the Tokyo Metropolitan Library. Anugraha Sinha walked us through the following:

Theory

1. Introduction to RL
2. Important elements of an RL problem
3. Description of Markov Decision Process (MDP) and and Markov Assumption.
4. Importance of parametrization of State, Action, Reward and Environment.
5. Model Based and Model Free Methods
6. Meaning of Control Problem and Evaluation Problem.
7. Algorithm of Policy Evaluation and Value iteration methods

Code examples

1. Finding the best route through a maze/obstruction avoidance using policy iteration algorithm.
2. Above problem statement with value iterations algorithm.
3. Code exercise

You can find the study material (slides and notebooks) for this session on GitHub. Join us for our next sessions on Meetup. We also provide a zoom link for participants to join remotely.