Adaptive resonance theory based neural network for supervised chemical pattern recognition (FuzzyARTMAP) Part 1: Theory and network properties

The FuzzyARTMAP algorithm is studied with respect to its usefulness for supervised chemical pattern recognition. The theory of this relatively complex artificial neural classifier is presented in detail for chemists. An instructive data set of moderate size, describing male and female participants in courses of chemometrics by their body measures, is used to demonstrate how FuzzyARTMAP works and what its basic properties are.