Reinforcement learning is a part of machine learning that tries to do things to maximize the rewards of actions in a particular environment. Reinforcement learning algorithms aim to find the best action in a given environment. This type of machine learning can learn to perform the learning process and achieve its goals even in complex and uncertain environments. Just like the human brain, the system is rewarded for good choices and penalized for bad decisions, and it learns from each action.
My principal reference for studying reinforcement learning is the Reinforcement Learning: An Introduction [1] book by Richard S. Sutton and Andrew Barto. I have also used the following courses to get started.
| Name of the course | Publisher's website | University or the instructor | The number of parts | |
|---|---|---|---|---|
| 1 | Reinforcement Learning Specialization | Coursera | University of Alberta | 4 |
| 2 | Reinforcement Learning with Pytorch | Udemy | Clive D. | 7 |
For more information, you can click on the course name you want.
In progress ....
This repository contains a deep reinforcement learning project that trains an agent to play the classic game of Pong. The algorithm used is based on the "Playing Atari with Deep Reinforcement Learning" article by the DeepMind team of Google. The project is implemented using Pytorch, numpy, matplotlib and gym.
The agent starts off making random moves, but as it learns from its experiences, it becomes better at playing the game and scores more points. The training process is accompanied by a collection of videos of the agent playing the Pong game, which can be found in the Pong Game Videos directory.
If you're interested in learning more about the project or replicating the training process, check out the Final Project directory for the code and results. And you can find more detailed information about this project in the repository of Artificial-Neural-Network
This repository is just a part of the main project which is focused on reinforcement learning and it's a great starting point for anyone interested in learning about deep reinforcement learning with Atari Pong.
This directory contains the materials for a workshop on Reinforcement Learning (RL), which covered a variety of topics including the fundamentals of RL, tabular methods, and the use of CNNs in RL. Additionally, attendees were able to participate in a hands-on demonstration of training an RL agent using the Pong game as an example. Resources such as slides and other reading materials are also included to further aid in understanding the topic.
[1] Richard S. Sutton and Andrew G. Barto. 2018. Reinforcement Learning: An Introduction. A Bradford Book, Cambridge, MA, USA.