Abstract
The initialisation of a neural network implementation of Sammon’s mapping, either randomly or based on the principal components (PCs) of the sample covariance matrix, is experimentally investigated. When PCs are employed, fewer experiments are needed and the network configuration can be set precisely without trial-and-error experimentation. Tested on five real-world databases, it is shown that very few PCs are required to achieve a shorter training period, lower mapping error and higher classification accuracy, compared with those based on random initialisation.
Similar content being viewed by others
Author information
Authors and Affiliations
Additional information
Received: 20 April 1999, Received in revised form: 08 July 1999, Accepted: 05 August 1999
Rights and permissions
About this article
Cite this article
Lerner, B., Guterman, H., Aladjem, M. et al. On the Initialisation of Sammon’s Nonlinear Mapping. Pattern Analysis & Applications 3, 61–68 (2000). https://doi.org/10.1007/s100440050006
Published:
Issue Date:
DOI: https://doi.org/10.1007/s100440050006