ARTIFICIAL REWARDS - A COMPARATIVE REVIEW OF REINFORCEMENT IN LEARNING-THEORY AND NEURAL COMPUTING