Model-based neural networks

We fitted a new method allowing to use some of the knowledge involved in ch emico-physical models with neural networks. This method, we call model-base d neural networks, needs the disposal of experimental measures, a fully det ermined model (even approximate), and access to the state variables of the system. Owing this, we are able to fundamentally include mathematical model s, such as physico-chemical ones, in the learning phase of the network, in order to improve its performances, although relying on experimental data. T he method was successfully tested by using theoretical examples. It occurs to be especially useful when experimental data are badly determined. (C) 20 01 Elsevier Science Ltd. All rights reserved.