ctRL

This project intends to be an introduction to reinforcement learning. It will involve the study and implementation of classic reinforment learning algorithms, one of which will be a simulation of the movement of a line following bot.

Table of contents

• About the project
• Getting started
• Future Work
• Resources

About the project

This project represents our journey in reinforcement learning. It includes the implementation of problems in reinforcement learning ranging from solving a simple version of the k-armed bandits problem to a simulation of the line following bot using Q-learning algorithm.

Getting started

Dependencies

• Python
• NumPy
• CoppeliaSim

Installations

Just clone this repository - git clone https://github.com/Ariv-Fernandes/ctRL_Eklavya23.git - and you're ready to go!

Future Work

1. Exploration and implementation of more problems covering the concepts of dynamic programming, Monte Carlo technique and Temporal-Difference learning.
2. We aim to implement the classic reinforcement learning algorithms, in which we’ll run it on an RPi to solve a control problem.

Resources

• Deep Reinforcement Learning
• Reinforcement Learning- Barto and Sutton
• Understand Q-Learning Algorithm