The standard Self-Organizing Map (SOM) achieves self organization through incremental, stepwise update of its weight vectors. However, beginning with a randomly initialized map, the convergence speed of SOM is rather slow. It would be desirable to speed up the self-organization process. For this reason, a new strategy for topological self-organization (Xu, 2002) has been proposed to speed up the self-organization process. This paper empirically compares the two strategies. The new strategy is shown to speed up the self-organization significantly. © Springer-Verlag 2003.
CITATION STYLE
Chan, B. K. Y., Chu, W. W. S., & Xu, L. (2004). Empirical comparison between two computational strategies for topological self-organization. Lecture Notes in Computer Science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2690, 410–414. https://doi.org/10.1007/978-3-540-45080-1_53
Mendeley helps you to discover research relevant for your work.