2 ms·
That one would be interesting to try with AlphaZero. Reinforcement learning systems tend to have trouble with imperfect information.
by sterlind 6y ago
That one would be interesting to try with AlphaZero. Reinforcement learning systems tend to have trouble with imperfect information.