Deep Reinforcement Learning

[jsbrinkman]

Thanks for the great work! Do you think that your work can also be applied to deep reinforcement learning? And if so, do you have an idea what would be the best option for this? Thank you in advance!