Learning and decision making in human during a game of matching pennies

摘要

This paper investigated the nature of learning process in human playing a competitive game with binary choices. As in reinforcement learning, the subject's choice during a competitive game was biased by its choice and reward history, as well as by the strategies of its opponent. Furthemore, consistent with predictions of the reinforcement learning model, we found that the magnitude of ERPs after losing to the computer opponent predicted whether subjects would change decision behavior on the subsequent trial, but in our experiment, the subject's decision wasn't systematically influenced by the cumulative effects of reward history.