Metrics and the Cooperative Process of the Self-organizing Map Algorithm
This paper explores effects of using different the distance measures in the cooperative process of the Self-Organizing Map algorithm on the resulting map. In standard implementations of the algorithm, Euclidean distance is normally used. However, experimentation with non-Euclidean metrics shows that this is not the only metric that works. For example, versions of the SOM algorithm using the Manhattan metric, and metrics in the same family as the Euclidean metric, can converge, producing sets of weight vectors indistinguishable from the regular SOM algorithm. However, just being a metric is not enough: two examples of such are described. Being analogous to the Euclidean metric is not enough either, and we exhibit members of a family of such distance measures that do not produce satisfactory maps.
KeywordsSelf-organizing map Metric Cooperative process
- 1.Haykin, S.: Neural Networks and Learning Machines. Prentice-Hall, New York (2009)Google Scholar
- 2.Jiang, F., Berry, H., Schoenauer, M.: The impact of network topology on self-organizing maps. In: Proceedings of First ACM/SIGEVO Summit on Genetic and Evolutionary Computation, GEC-2009, pp. 247–254. ACM, New York (2009)Google Scholar
- 5.Płoński, P., Zaremba, K.: Improving performance of self-organising maps with distance metric learning method. In: Rutkowski, L., Korytkowski, M., Scherer, R., Tadeusiewicz, R., Zadeh, L.A., Zurada, J.M. (eds.) ICAISC 2012. LNCS, vol. 7267, pp. 169–177. Springer, Heidelberg (2012). doi: 10.1007/978-3-642-29347-4_20 CrossRefGoogle Scholar