evoNN_UTTT v0.1
Introduction It’s 2026. I shouldn’t need to exert myself having fun playing board games when a computer can have fun for me. Enter: evoNN-uttt — a neural network trained to beat me at the game of ‘Ultimate Tic-Tac-Toe’. Aside from some brief experiments conducted a few months ago (consisting of optimising hyperparameters for a genetic algorithm that solved the pure knapsack problem), I do not have much experience with evolutionary algorithms (note: I will use the terms evolutionary algorithm and genetic algorithm interchangeably), so iterating on and optimising a library for running the algorithm efficiently should be an interesting learning experience. ...