A deep Q-network solution of the CartPole problem, prepared for the Collabor8.1 conference.
This project requires Python 3 and Tensorflow to work. It provides a framework for solving reinforcement learning problems using deep Q-networks.
Much of the code is inspired by other implementations. In particular, the hyperparameter choices are based on YADQN.