Of Q‑Tables and Three‑in‑a‑Rows: Training an RL Knight in Tic‑Tac‑Toe
Reinforcement Learning is fairly popular at the moment. In this chronicle, we shall embark on a quest to forge a Reinforcement Learning model for the noble game of Tic‑Tac‑Toe. We’ll write our own environment, summon a DQN sorcerer from Stable Baselines 3, and ultimately witness our AI crush the humblest of human challengers (or at least draw more than half the time).