A Novel Ensemble of Scale-Invariant Feature Maps
A novel method for improving the training of some topology preserving algorithms as the Scale Invariant Feature Map (SIM) and the Maximum Likelihood Hebbian Learning Scale Invariant Map (MAX-SIM) is presented and analyzed in this study. It is called Weighted Voting Superposition (WeVoS), providing two new versions, the WeVoS-SIM and the WeVoS-MAX-SIM. The method is based on the training of an ensemble of networks and the combination of them to obtain a single one, including the best features of each one of the networks in the ensemble. To accomplish this combination, a weighted voting process takes place between the units of the maps in the ensemble in order to determine the characteristics of the units of the resulting map. For comparison purposes these new models are compared with their original models, the SIM and MAX-SIM. The models are tested under the frame of an artificial data set. Three quality measures have been applied for each model in order to present a complete study of their capabilities. The results obtained confirm that the novel models presented in this study based on the application of WeVoS can outperform the classic models in terms of organization of the presented information.
Unable to display preview. Download preview PDF.
- 5.Corchado, E., Fyfe, C.: Maximum Likelihood Topology Preserving Algorithms. In: U.K. Workshop on Computational Intelligence (2002)Google Scholar
- 6.Corchado, E., Fyfe, C.: The Scale Invariant Map and Maximum Likelihood Hebbian Learning. In: International Conference on Knowledge-Based & Intelligent Information & Engineering System (2002)Google Scholar
- 9.Fyfe, C., Corchado, E.: Maximum likelihood Hebbian rules. In: European Symposium on Artificial Neural Networks (ESANN) (2002)Google Scholar
- 10.Georgakis, A., Li, H., Gordan, M.: An ensemble of SOM networks for document organization and retrieval. In: International Conference on Adaptive Knowledge Representation and Reasoning (AKRR 2005), pp. 6–141 (2005)Google Scholar
- 11.Heskes, T.: Balancing between bagging and bumping. In: Advances in Neural Information Processing Systems, vol. 9, pp. 466–472 (1997)Google Scholar
- 12.Johansson, U., Lofstrom, T., Niklasson, L.: Obtaining accurate neural network ensembles. In: International Conference on Computational Intelligence for Modelling, Control & Automation Jointly with International Conference on Intelligent Agents, Web Technologies & Internet Commerce, Proceedings, vol. 2, pp. 103–108.Google Scholar
- 14.Kiviluoto, K.: Topology preservation in self-organizing maps. In: IEEE International Conference on Neural Networks (ICNN 1996), vol. 1, pp. 294–299 (1996)Google Scholar
- 19.Petrakieva, L., Fyfe, C.: Bagging and Bumping Self Organising Maps. Computing and Information Systems Journal, 1352–1404 (2003)Google Scholar
- 20.Polani, D.: Measures for the Organization of Self-Organizing Maps. In: Self-organizing Neural Networks: Recent Advances and Applications. Studies in Fuzziness and Soft Computing, pp. 13–44 (2003)Google Scholar
- 21.Polzlbauer, G.: Survey and Comparison of Quality Measures for Self-Organizing Maps. In: Fifth Workshop on Data Analysis (WDA 2004), pp. 67–82. Elfa Academic Press, London (2004)Google Scholar
- 23.Vesanto, J., Sulkava, M., Hollen, J.: On the Decomposition of the Self-Organizing Map Distortion Measure. In: Proceedings of the Workshop on Self-Organizing Maps (WSOM 2003), pp. 11–16 (2003)Google Scholar
- 24.Voronoi, G.: Nouvelles applications des parametres continus a la theorie des formes quadratiques. Math. 133, 97–178 (1907)Google Scholar