How training and testing histories affect generalization: a test of simple neural networks

Abstract

We show that a simple network model of associative learning can reproduce three findings that arise from particular training and testing procedures in generalization experiments: the effect of 1) ``errorless learning'' and 2) extinction testing on peak shift, and 3) the central tendency effect. These findings provide a true test of the network model, which was developed to account for other penhomena, and highlight the potential of neural networks to study phenomena that depend on sequences of experiences with many stimuli. Our results suggest that at least some such phenomena, e.g., stimulus range effects, may derive from basic mechanisms of associative memory rather than from more complex memory processes.