4 ms·
There is a tic-tac-toe system that learns to play using reinforcement learning built of matchboxes. Designed in 1961. https://en.wikipedia.org/wiki/Matchbox_Edu
by fhars 3y ago
There is a tic-tac-toe system that learns to play using reinforcement learning built of matchboxes. Designed in 1961. https://en.wikipedia.org/wiki/Matchbox_Educable_Noughts_and_Crosses_Engine https://en.wikipedia.org/wiki/Matchbox_Educable_Noughts_and_...
- ufocia 3y agoInteresting. Only 304 unique permutations so 9 bits is more than anyone would ever need for a tic-tac-toe state?