Deep Q Network RL Agent
This repository is based off of (https://github.com/matthiasplappert/keras-rl).
New exploration and exploitation strategies have been implemented on top of the base code.
Run the Code
The example files detail how to set up and run a reinforcement learning agent using Q-learning with a convolutional neural network for processing the raw pixels from ATARI games.
Dependencies
- Tensorflow
- Open AI Gym
- Keras
The above packages require additional dependencies, see their documentation for more information.