Asynchronous Methods for Deep Reinforcement Learning

https://arxiv.org/pdf/1602.01783.pdf