Deep reinforcement learning

For details, have a look at the CMS.