Internal Categories with Irregular Geometry and Overlapping in ART Networks
PolyTope ARTMAP (PTAM)  is an ART neural network based on internal categories with irregular polytope (polygon in IR n ) geometry. Categories in PTAM do not overlap, so that their expansion is limited by the other categories, and not by the category size. This makes the vigilance parameter unnecessary. What happens if categories have irregular geometries but overlap is allowed? This paper presents Overlapping PTAM (OPTAM), an alternative to PTAM based on polytope overlapping categories, which tries to answer this question. The comparison between the two approaches in classification tasks shows that category overlap does not reduce neither the classification error nor the number of categories, and it also requires vigilance as a tuning parameter. Futhermore, OPTAM provides a significant variability in the results among different data sets.
KeywordsInternal Category Input Pattern Category Size Output Prediction Irregular Geometry
Unable to display preview. Download preview PDF.
- 1.Anagnostopoulos, G., Georgiopoulos, M.: Ellipsoid ART and Ellipsoid ARTMAP for incremental clustering and classification. IEEE Int. Joint Conf. on Neural Networks 2, 1221–1226 (2001)Google Scholar
- 5.Gomes, D., Fernández-Delgado, M., Barro, S.: Simplex ARTMAP: building general geometry borders among predictions with simplex-shaped classes. IASTED Int. Conf. on Artificial Intelligence and Soft Computing 1, 232–237 (2004)Google Scholar
- 6.Gomes, D., Fernández-Delgado, M., Barro, S.: A vigilance-free ART network with general geometry internal categories. In: 2005 IEEE Int. Joint Conf. on Neural Networks, pp. 463–468 (2005)Google Scholar
- 7.Gómez-Sánchez, E., Dimitriadis, Y., Cano-Izquierdo, J., Coronado, J.: MicroARTMAP: use of mutual information for category reduction in Fuzzy ARTMAP. In: IEEE Int. Joint Conf. on Neural Networks, pp. 647–652 (2000)Google Scholar
- 10.Wilensky, G.: Analysis of neural network issues: scaling, enhanced nodal processing, comparison with standard classification. DARPA Neural Network Program Review (1990)Google Scholar