AI
2002
Springer
13 years 11 months ago
2002
Springer
TD-Gammon is a neural network that is able to teach itself to play backgammon solely by playing against itself and learning from the results. Starting from random initial play, TD...