On March 9-15, 2016, the AlphaGo program, created by Google DeepMind based on the methods of deep reinforcement learning, won Li Sedol, a GO 9 professional and one of the best human players, with a score of 4-1, and most recently AlphaZero has learned to play GO, chess and shogi are better than the former, without using any outside information at all, just the rules of the game. In the report, we will try to answer the following questions:
– why it is so important and difficult, after all, it would seem, DeepBlue beat Kasparov ten years
– What is deep reinforcement learning, how does it work?
– What are the main ideas of AlphaGo actually, what is the breakthrough?
– why these toys? why can AlphaGo ideas be used in particular and deep reinforcement learning in general?
Buy this talk
Buy this video
ConferenceCast.tv — conference video talk archive.
With this service you can find interesting talks especially for you!