April 2018
Intermediate to advanced
334 pages
10h 18m
English
Go cannot be approached in a game tree way. The reason is that the bigger complexity and brute force approach used don't perform any sort of learning. The only task it performs is mapping the game state to a node in the game tree. Moreover, the brute force approach used in DeepBlue didn't have a generalized evaluation function rather it was hand-crafted for different game positions. Thus the preceding approach is too game specific and such approaches can't be scaled up to play Go.