A unified framework for knowledge based spectrum analysis

A unified approach for the development of a knowledge-based spectral analysis system is presented. Most knowledge available in the literature is based on the comparison of one method with a few competing methods. This problem is offset by setting up performing indexes on a common framework. The importance of using psychometric tools during knowledge acquisition is discussed, and the repertory grid and ADCLUS (additive clustering) model are shown to yield adequate results. A framework for representing knowledge in spectral analysis is provided. The use strategy provides an environment where spectral analysis methodological research can be constructed. >