D. Wienke et L. Buydens, ADAPTIVE RESONANCE THEORY-BASED NEURAL NETWORKS - THE ART OF REAL-TIME PATTERN-RECOGNITION IN CHEMICAL PROCESS MONITORING, TrAC. Trends in analytical chemistry, 14(8), 1995, pp. 398-406
The family of artificial neural networks based on Adaptive Resonance T
heory (ART) forms a collection of distinct mathematical pattern-recogn
ition methods, The classification of sensor signals, process data anal
ysis, spectral interpretation, and image analysis are discussed as app
lications of ART outside and within chemistry, The advantages of ART a
re considered. They include its use as a built-in detector for outlier
s, its rapid training speed, self-organizational behaviour, full chemi
cal interpretability, and real-time and on-line applicability. A gloss
ary of terms used in ART is given at the end of the article.