D-ART: a pattern recognition system based on adaptive resonance and algebraic metric space theories

Summary form only given, as follows. The authors discuss combining the adaptive resonance theory of neural networks with results from algebraic metric space theory. The notion of the so-called dual planes is adapted to Carpenter and Grossberg´s adaptive resonance theory systems (hence the name D-ART). The new system is appropriate for a large class of applications, for instance, `sparse´ images, e.g., wire-frame or constellation data. Unlike the classical ART, D-ART explicitly utilizes the geometry of the underlying space