THE STATISTICAL-MECHANICS OF LEARNING A RULE

A summary is presented of the statistical mechanical theory of learnin g a rule with a neural network, a rapidly advancing area which is clos ely related to other inverse problems frequently encountered by physic ists. By emphasizing the relationship between neural networks and stro ngly interacting physical systems, such as spin glasses, the authors s how how learning theory has provided a workshop in which to develop ne w, exact analytical techniques.