Deep Reinforcement Learning for Doom using Unsupervised Auxiliary Tasks | Read Paper on Bytez