Arousal-performance Interactions In neural networks: The Yerkes-Dodson law revisited

摘要

A neural network model of the Yerkes-Dodson Law is described. The network's leaming performance varies as a function of simulated arousal and of task difficulty, in the way dscribed by the Yerkes-Dodson Law: The arousal-performance relationship is of an inverted-U form and optimal arousal is higher for easier tasks.