Recent Posts

Welcome to my blog.

Visualising Four Dimensions

27 October 2019

Temporal Difference Learning For Ultimate Tic Tac Toe

19 September 2019

This post describes the implementation of temporal difference learning that can be found on my github. This amazingly simple algorithm is able to learn entirely through self-play without any human knowledge, except for the rules of the game. By way of example, we will be training the algorithm to play ultimate tic-tac-toe, which I have already discussed here, but the same algorithm can be applied to almost any other game with varying degrees of success. This post will assume some familiarity with machine learning and reinforcement learning concepts, and should be accessible if you understand the basics of supervised learning with neural networks.

Artificial Intelligence And Ultimate Tic Tac Toe

18 September 2019

You probably already know how to play the original version of tic-tac-toe. Well ultimate tic-tac-toe is similar, except that each square on the board contains another, smaller game of tic-tac-toe! Let’s call the big game the “macro game” and the small games “micro games”. A player must win micro games to claim corresponding squares in the macro game. The goal is to win three micro games in a row. Simple, right?

Kelsey Is Cute

17 September 2019

ultimate tic-tac-toe gif

Figure 1: Kelsey being cute.