The use case is similar to AlphaGo, ie deterministic full information game with 
a huge state space.

That reinforcement package looks interesting, looking into it. Love to 
collaborate

Reply via email to