Research on some problems in the Kohonen SOM algorithm

Author

He, Ying ; Feng, Tian-Jin ; Cao, Jun-Kuo ; Ding, Xiang-Qian ; Zhou, Ying-Hui

Author_Institution

Dept. of Electr. Eng., Ocean Univ. of Qingdao, China

Volume

3

fYear

2002

fDate

2002

Firstpage

1279

Abstract

The article analyzes the relation between initial parameters setting and the formation of a topographic map of the input patterns in which the spatial locations of the neurons in the lattice are indicative of intrinsic statistical features contained in the input patterns of a Kohonen self-organizing map (SOM) algorithm. Taking a network arranged in the form of a two-dimensional lattice and trained with a two-dimensional input vector as an example, the author puts forward an initializing method for connection weights of the neurons in the competition layer.

Keywords

learning (artificial intelligence); self-organising feature maps; Kohonen SOM algorithm; Kohonen self-organizing map algorithm; competition layer; connection weights; initial parameters setting; initializing method; input patterns; intrinsic statistical features; topographic map; two-dimensional input vector; two-dimensional lattice; Application software; Convergence; Helium; Lattices; Machine learning; Machine learning algorithms; Neurons; Oceans; Organizing; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Machine Learning and Cybernetics, 2002. Proceedings. 2002 International Conference on

Print_ISBN

0-7803-7508-4

Type

conf

DOI

10.1109/ICMLC.2002.1167409

Filename

1167409