Title :
Reducing generalization error and category proliferation in ellipsoid ARTMAP via tunable misclassification error tolerance: boosted ellipsoid ARTMAP
Author :
Anagnostopoulos, Georgios C. ; Georgiopoulos, Michael ; Verzi, Stephen J. ; Heileman, Gregory L.
Author_Institution :
Univ. of Central Florida, Orlando, FL, USA
fDate :
6/24/1905 12:00:00 AM
Abstract :
In this paper we introduce boosted ellipsoid ARTMAP (bEAM), a variant of ellipsoid ARTMAP, which via a tunable misclassification error tolerance increases the network´s resubstitution error and implicitly performs structural risk minimization. bEAM constitutes another example of how modifications to fuzzy ARTMAP can be naturally extended to ellipsoid ARTMAP
Keywords :
ART neural nets; fuzzy neural nets; generalisation (artificial intelligence); minimisation; bEAM; boosted ellipsoid ARTMAP; category proliferation reduction; ellipsoid ARTMAP; fuzzy ARTMAP; generalization error reduction; resubstitution error; structural risk minimization; tunable misclassification error tolerance; Ellipsoids; Fuzzy logic; Geometry; Neural networks; Resonance; Risk management; Size measurement; Subspace constraints; Testing; Training data;
Conference_Titel :
Neural Networks, 2002. IJCNN '02. Proceedings of the 2002 International Joint Conference on
Conference_Location :
Honolulu, HI
Print_ISBN :
0-7803-7278-6
DOI :
10.1109/IJCNN.2002.1007562
