Abstract
The recent Polytope ARTMAP (PTAM) suggests that irregular polytopes are more flexible than the redefined category geometries to approximate the borders among the desired output predictions. However, category expansion and adjustment steps without statistical information make PTAM not robust to noise and category overlap. In order to push the learning problem towards Structural Risk Minimization (SRM), this paper proposes Hierarchical Polytope ARTMAP (HPTAM) to use a hierarchical structure with different levels, which are determined by the complexity of regions incorporating the input pattern. Besides, overlapping of simplexes from the same desired prediction is designed to reduce category proliferation. Although HPTAM is still inevitably sensible to noisy outliers in the presence of noise, main experimental results show that HPTAM can achieve a balance between representation error and approximation error, which ameliorates the overall generalization capabilities.
Similar content being viewed by others
References
Verzi S J, Heileman G L. Generalization performance of ARTMAP-based networks in structural risk minimization framework. In Proc. Int. Joint Conf. Neural Netw., Hawaii, USA, May 12-17, 2002, pp.2644–2649.
Amorim D G, Delgado M F, Ameneiro S B. Polytope ARTMAP: Pattern classification without vigilance based on general geometry categories. IEEE Trans. Neural Netw., 2007, 18(5): 1306–1325.
Hush D. Learning from examples: From theory to practice. In ICNN 97 Tutorial, Huston, USA, Jun. 1997, Notes/Viewgraphs.
Williamson J R. Gaussian ARTMAP: A neural network for fast incremental learning of noisy multidimensional maps. Neural Netw., 1996, 9(5): 881–897.
Anagnostopoulos G C, Georgiopoulos M. Hypersphere ART and ARTMAP for unsupervised and supervised incremental classification. In Proc. Int. Joint Conf. Neural Netw., Como, Italy, Jul. 24-27, 2000, pp.59–64.
Anagnostopoulos G C, Georgiopoulos M. Ellipsoid ART and Ellipsoid ARTMAP for incremental clustering and classification. In Proc. Int. Joint Conf. Neural Netw., Washington DC, USA, 2001, pp.1221–1226.
Carpenter G A, Ross W D. ART-EMAP: A neural network architecture for object recognition by evidence accumulation. IEEE Trans. Neural Netw., 1995, 6(4): 805–818.
Carpenter G A, Milenova B L, Noeske B W. Distributed ARTMAP: A neural network for fast distributed supervised learning. Neural Netw., 1998, 11(5): 793–813.
Liao L, Wu Y. Distributed Polytope ARTMAP: A vigilance-free ART network for distributed supervised learning. In Proc. Int. Joint Conf. Computational Sciences and Optimization (CSO), Hong Kong, China, Apr. 24-26, 2009, pp.501–504.
Verzi S J, Heileman G L, Georgiopoulos M, Healy M J. Boosting the performance of ARTMAP. In Proc. Int. Joint Conf. Neural Netw., Anchorage, USA, May 4-9, 1998, pp.369–401.
Gómez-Sánchez E, Dimitriadis Y A, Cano Izquierdo J M, López-Coronado J. μARTMAP: Use of mutual information for category reduction in Fuzzy ARTMAP. IEEE Trans. Neural Netw., 2002, 13(1): 58–69.
Verzi S J, Heileman G L, Georgiopoulos M, Healy M J. Hierarchical ARTMAP. In Proc. Int. Joint Conf. Neural Netw., Como, Italy, Jul. 24-27, 2000, pp.41–46.
Verzi S J, Heileman G L, Georgiopoulos M, Healy M J. Rademacher penalization applied to fuzzy ARTMAP and boosted ARTMAP. In Proc. Int. Joint Conf. Neural Netw., Washington DC, USA, 2001, pp.1191–1196.
Marriott S, Harrison R F. A modified fuzzy ARTMAP architecture for the approximation of noisy mappings. Neural Netw., 1995, 8(4): 619–641.
Healy M J, Caudell T P, Smith S D G. A neural architecture for pattern sequence verification through inferencing. IEEE Trans. Neural Netw., 1993, 4(1): 9–20.
Cano Izquierdo J M, Dimitriadis Y A, Gómez Sánchez E, López Coronado J. Learning from noisy information in FasArt and FasBack neuron-fuzzy systems. Neural Netw., 2001, 14: 407–425.
Author information
Authors and Affiliations
Corresponding author
Additional information
Supported by the National Basic Research 973 Program of China under Grant No. 2007CB311006.
Rights and permissions
About this article
Cite this article
Liao, L., Wu, YQ. & Han, CZ. Hierarchical Polytope ARTMAP for Supervised Learning. J. Comput. Sci. Technol. 25, 1071–1082 (2010). https://doi.org/10.1007/s11390-010-9388-z
Received:
Revised:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s11390-010-9388-z