Mini-Alpha-Go

From scratch implementation of a 9x9 go game engine and elo-tracker, alongside an Alpha-Go implementation that uses Monte-Carlo tree search wth deep residual networks to make its next move, trained using combination of self play and supervised learning.