This is the blog for a project that I have started after attending the fast.ai online course.

The goal is to use deep learning to explore the dynamics of a simple paper-and-paper game called Morpion Solitaire.

gif

My approach is presented in a series of blog posts:

  1. The game: What is the game about? Why is this an interesting framework for deep learning?
  2. The data: What is the data used for deep learning? How is it organized?
  3. A simple binary problem: Answering a yes/no question with a basic model.
  4. Going deeper with a counting problem: Solving a more complicated problem with a deeper network.
  5. Playing with a model: How do we use the model to improve our exploration of the game.
  6. My first ResNet (coming up): Starting to make real predictions with a residual network.

This is on-going work, so stay tuned for updates!

Posts

subscribe via RSS