Monte Carlo tree search

When we talk about games such as chess, Go, or Tic-Tac-Toe, which are games of strategy based upon the current scenario, we're talking about a huge number of possible scenarios and actions that can be performed in them at any given point. While, for smaller games such as tic-tac-toe, the number of possible states and actions is within the reach of modern-day computers to calculate, more complex games such as chess and Go exponentially increase in terms of how many possible states of the game can be made.

The Monte Carlo tree search attempts to find the right series of actions required to win any game or to achieve a better reward in a given environment. The reason it is called a tree search is due to the fact that ...

Get Mobile Deep Learning with TensorFlow Lite, ML Kit and Flutter now with the O’Reilly learning platform.

O’Reilly members experience books, live events, courses curated by job role, and more from O’Reilly and nearly 200 top publishers.