As far as I can tell (correct me if I'm wrong), Alphazero (with MCTS and neural network heuristic function RL) is the state of the art training method for turn based, deterministic, perfect information, complete information, two player, zero sum games.
But what is the state of the art for turn based, imperfect information games, that have 2 players, complete information, and is zero sum? (Deterministic or stochastic.) Examples include Battleship and most 2 player card games.
Are there standard games, or other tests by which this is measured? Is the criteria I offered for type of game not specific enough to narrow the answer down properly?
If the state of the art involves supervised learning (data set of manually played games), then what's the state of the art for pure reinforcement learning, if there is one?