An introduction to Deep Q-Learning: let's play Doom