← Back to Projects

Connect Four

Monte Carlo Tree Search
Your Turn (Click a column)
200

Unlock this to tune and understand how search depth impact the algorithm strength performance.

How it works

The game is the implementation of Connect Four where the opponent is the Monte Carlo Tree Search (MCTS) algorithm or, if selected, the Alpha MCTS. This algorithm is known for its ability to compute in near real time the best next move based on a large optimized computation of options.

Connect Four State Representation

The Image shows that the q-values are not exactly the probability of winning, but the expected reward from taking that action. The selected action is the one with the highest q-value.

Algorithm Steps

todo//

  • Input: todo//
  • Job: todo//
  • Output: todo//

todo//

Backend Implementation

The backend running the neural network is decoupled from the website. For each user action, the backend is invoked via an API call to an external endpoint hosted on AWS. The backend is deployed as an AWS Lambda function running a FastAPI application packaged as an ECR container image.

AWS Lambda Implementation

References & Code

The algorithm and training code running inside the AWS Lambda function is available on GitHub: