RL workshop tutorial

Repo contributors:

Zeming Fang, zemingfang11@gmail.com

Ru-Yuan Zhang, ruyuanzhang@gmail.com

Jieying Zhang, jieying.zhang@student.uva.nl

Getting started

Install Anaconda for later python installation. Open the Terminal (for windows user, please use git bash).

Create a virtual environment

conda create --name rlworkshop python=3.10

Activate the environment

conda activate rlworkshop

Install the dependencies

pip install -r requirements.txt

TOC

1. Intro to RL

   • Frozen lake environment
   • Policy evaluation

2. Dynamic Programming and Monte Carlo Methods

   • Policy improvment
   • Value iteration
   • Monte Carlo prediction

3. Temporal Difference Methods

   • Monte Carlo control
   • Sarsa
   • Q-learning