Neurospectra-a neural network model for spectral estimation

An artificial neural network (ANN) model for spectral estimation is presented. The objective is to reduce the lengthy knowledge acquisition process in expert systems by using the supervised learning process of the ANN. The design of the spectral estimation system and the corresponding training system are presented. The implementation of the spectral estimator is realized using programs in conventional programming languages for the numeric routines (spectral estimators, model validation, etc.) and Prolog for the ANN module.<
>