You are viewing a single comment's thread from:
RE: Artificial intelligence for Dummies: It's All In The Game (Part 4)
This is great! On chess and artificial intelligence, the Kasparov vs Deep Blue match was epic, but projects like AlphaZero and Leela are even more impressive. The Monte Carlo algorithm seems to be better than MinMax in this field.
Hi @eniolw, thanks :))
The Monte Carlo can be extremely useful in cases when we don't have fully observable environment like when we play cards or when we have some stochastic process like throwing dice :)
Minimax was just an example how this logic works, I personally don't know is it still used somewhere today:)
It is! conventional chess engines use extensively MinMax as search algorithm, but in December of last year, Google proved with AlphaZero that Monte Carlo and reinforcement learning can achieve better results. That is why Leela Chess was born and currently she has caught top chess engines in playing strength.