cassincruz / connect4-ai

Experimenting and exploring Reinforcement Learning by attempting to create a Connect4 AI.

Connect 4 Reinforcement Learning

Key files and folders

• Connect4.py: Defines a custom Connect4 Gymasium environment to be used during training.
• Connect4DQN.py: Defines agent functions using Deep Q Learning and a neural network architecture.
• nn.py: Used to define and import PyTorch neural network architectures.
• Training.ipynb: Jupyter notebook used for training agents and visualizing weights.

References & Helpful Resources

• Machine Learning with Phil - "Deep Q Learning is Simple with PyTorch | Full Tutorial 2020" ^
• Alberto Bas - Training a DQN agent to play Connect Four ^
• Codebox Software (Rob Dawson) - Learning to play Connect 4 with Deep Reinforcement Learning ^
• Calin - Deep Learning Architectures, A Mathematical Approach