Build algorithm based on AlphaZero which uses Monte Carlo Tree Search combined with neural network based policy and value function to play two player zero sum games. Test algorithm based on game of Chopsticks
Modified code from https://github.com/JoshVarty/AlphaZeroSimple/