SYSTEMATIC CLASSIFICATION AND IDENTIFICATION OF NOISE SPECTRA USING PERCEPTRON-BASED NEURAL NETWORKS

A general framework for the detectin of gradually developing changes i n a noise generating system is presented. The procedure is based on a new learning algorithm developed for neural networks with dynamically building architecture. The method has been tested by using almost a th ousand noise spectra recorded from different detector types and from d ifferent detector positions. This work is part of a larger project, ai med at developing a noise diagnostic expert system.