Human-level control through deep reinforcement learning

[martin]

http://www.nature.com/nature/journal/v518/n7540/full/nature14236.html