Studies in machine cognition using the game of poker

Abstract

A progress report is presented of on-going research efforts concerning human decision making under uncertainty and risk and human problem solving and learning processes on the one hand, and machine learning, large scale programming systems, and novel programming techniques on the other. There has also been interest in how humans make deductive and inductive inferences and form and optimize heuristic rules, and how machines can reach similar results. Although the vehicle of these investigations has been the game of poker, a conceptual framework has been provided that should have a fairly wide range of applicability. The models of human judgement, choice, and decision making are incorporated in a large scale complex program. They represent both descriptive and normative theories of behavior. An interactive game environment has been recently established which, besides its usefulness for experiments in game playing, enables humans to construct machine strategies “on-line” in a question answering, advice taking mode.