DeckOfCards

This project involves designing environments to simulate various card games in order to test out various reinforcement learning algorithms. The primary goals are as follows:

1. Through self-play, train agents capable of outperforming humans
2. Train agents of varying skill levels
3. Train agents across a variety of domains (card games)
4. Learn approaches that can generalize across domains and can be potentially applied to other domains (e.g. AI in video games)

Motivation

Reinforcement learning is a field of research that has enormous potential. Two of the biggest difficulties with reinforcement learning is that it requires a significant amount of training time and does not generalize well across domains. This project aims to study reinforcement learning in order to better understand applying it to different domains and determine if either of these problems can be solved.

A second motivation I have is that typically playing against AI in games can be extremely predictable. Due to the rule-based approach to developing AI bots, playing against them can be uninteresting in the long run. Developing a generalizable approach to training AI to play in a more interesting manner would change the industry.

Features

TBD

Getting Started

1. Fork the project
2. Clone the repository: $git clone https://github.com/Atrus619/DeckOfCards
3. Create a virtual environment: $python3 -m venv /path/to/new/virtual/env
4. Install dependencies: $pip install -r requirements.txt

How to Use?

TBD

How to Contribute

TBD

Credits

TBD

License

MIT License