Poker pro and software developer Nikolai Yakovenko continues his three-part series examining how far researchers have gotten in their efforts to build a hold’em playing AI system.
According to my buddy Justin, classification and its cousin regression are essentially a science, while clustering is more of an art.
Many of us are familiar with classification and regression problems. Will the dollar go up or down against the euro? Which horse should I back in the next tournament? Is it better to raise, call, or fold in this position, or should I mix it up? There is an objective that you’d like to maximize, and there is a science to maximizing that objective, even if we might disagree what that objective should be.