Neural network model of serial learning

Abstract

A brief introduction to neural networks is given, and correspondences between the elements of such networks and the primitive objects of APL are indicated. A neural network model is presented for simulating serial learning, the ability of an animal to respond in a prescribed sequence to an array of stimuli presented simultaneously. The particular network presented is intended to model stimulus-response (s-r) chaining, a behaviorist theory of serial learning. Some results obtained from running the network are discussed.