Simple Reinforcement Learning with Tensorflow Part 0: Q-Learning with Tables and Neural Networks
Arthur Juliani
7.7K76
Great inspiring stuff. I have now tried the Cart Pole in OpenAI gym, with Keras Nueral Nets.
I have also used a Keras based DQN Implementation for a Pong Playing Game. [Not a convolutional Net, just basic paddle and ball state] I built Pong game against PyGame as the OpenAI Atari examples don’t run under Windows.