Absolutely. Q-learning has this capabilities and a shallow neural network was used back in 1992 to play backgammon, which has a lot of stochasticity. See https

Absolutely. Q-learning has this capabilities and a shallow neural network was used back in 1992 to play backgammon, which has a lot of stochasticity. See https://en.wikipedia.org/wiki/TD-Gammon https://en.wikipedia.org/wiki/TD-Gammon