On Growing Self - Organizing Neural Networks without Fixed Dimensionality

  • Authors:

  • Guojian Cheng;Ziqi Song;Jinquan Yang;Rongfang Gao

  • Affiliations:

  • Xi'an Shiyou University Shaanxi Province, 710065, China;Xi'an Shiyou University Shaanxi Province, 710065, China;Xi'an Shiyou University Shaanxi Province, 710065, China;Xi'an Shiyou University Shaanxi Province, 710065, China

  • Venue:
  • CIMCA '06 Proceedings of the International Conference on Computational Inteligence for Modelling Control and Automation and International Conference on Intelligent Agents Web Technologies and International Commerce
  • Year:
  • 2006

Quantified Score

Hi-index 0.00

Visualization

Abstract

Kohonen's Self-Organizing Maps (KSOM) can generate mappings from high-dimensional signal spaces to lower dimensional topological structures. The main features of this kind of mappings are formation of topology preserving, feature mappings and probability distribution approximation of input patterns. However, KSOM have some limitations, e.g., a fixed number of neural units and a topology of fixed dimensionality, which makes KSOM impractical for applications where the optimal number of units is not known in advance and resulting in problems if this predefined dimensionality does not match the dimensionality of the feature manifold. Growing Self-Organizing Neural Networks (GSONN) can change their topological structures during learning. GSONN without fixed dimensionality has no topology of a fixed dimensionality imposed on the network. This paper first gives an introduction to neural gas network, a non-grid KSOM. Then, we discuss some GSONN without fixed dimensionality such as growing neural gas and the author's model: twin growing neural gas. It is ended with some testing results comparison and conclusions.