Artificial neural networks as models of stimulus control

We evaluate the ability of artificial neural network models (multilayer per ceptrons) to predict stimulus-response relationships. A variety of empirica l results are considered, such as generalization, peak shift (supernormalit y) and stimulus intensity effects. The networks were trained on the same ta sks as the animals in the experiments considered. The subsequent generaliza tion tests on the networks showed that the model replicates correctly the e mpirical results. We conclude that these models are valuable tools in the s tudy of animal behaviour. (C) 1998 The Association for the Study of Animal Behaviour.