You should check out Berkeley's deep reinforcement learning course[1]. There's lecture videos, slides, and homework assignments, and it's all very up-to-date.

[1] http://rll.berkeley.edu/deeprlcourse/