Abstract

A performance comparison of two self-organizing networks, the Kohonen Feature Map and the recently proposed Growing Cell Structures is made. For this purpose several performance criteria for self-organizing networks are proposed and motivated. The models are tested with three example problems of increasing difficulty. The Kohonen Feature Map demonstrates slightly superior results only for the simplest problem. For the other more difficult and also more realistic problems the Growing Cell Structures exhibit significantly better performance by every criterion. Additional advantages of the new model are that all parameters are constant over time and that size as well as structure of the network are determined automatically. 1 INTRODUCTION Self-organizing networks are able to generate interesting low-dimensional representations of high-dimensional input data. The most well-known of these models is the Kohonen Feature Map (Kohonen [1982]). So far it has been applied to a large variety of ...

Citations

228	Growing Cell Structures - A SelfOrganizing Network for Unsupervised and - Fritzke - 1994
1	Pawelzik [1992], "Quantifying the neighborhood preservation of self-organizing feature maps - Bauer
1	1991], "A study of the application of Kohonen-type neural networks to the travelling Salesman Problem - Favata
1	1991], "Unsupervised clustering with growing cell structures - Fritzke
1	1991], "FLEXMAP - A neural network with linear time and space complexity for the traveling salesman problem - Fritzke
1	1982], "Self-Organized Formation of Topologically Correct Feature Maps - Kohonen
1	Darken [1989], "Fast Learning - Moody
1	Schulten [1990], "Large-scale simulations of selforganizing neural networks on parallel computers: application to biological modeling - Obermayer, Ritter
1	Kohonen [1989], "Self-Organizing Semantic Maps - Ritter