MULTIPARAMETER PROCESS PREDICTION WITH NEURAL NETWORKS

Neural networks (NNs) are used to predict characteristics of GaAlAs la yers grown by organometallic chemical vapor deposition (OMCVD). Tradit ional statistical techniques fail because there are many parameters wh ich control the growth process and relatively few experiments to allow a full description of the effect of changing parameters. A successive approximation technique with NNs was developed which enables the most relevant input parameters to be selected first by a linear NN and the n used by a more general NN to accurately predict the layer characters itics. In addition, by training to predict the correction to analytic approximations for the layer characteristics, maximum use is made of p rior knowledge about the problem which results in a significant improv ement in predictive capability beyond the simple analytic approximatio ns.