ANU Computer Science Technical Reports
TR-CS-12-02
Dingyun Zhu Huajie Wu, Tom Gedeon.
Spherical Topology Self-Organazing Map Neuron Network for
Visualization of Complex Data.
February 2012.
[POSTSCRIPT (4185240 bytes)] [PDF (1621816 bytes)]
Abstract: The spherical SOM (SSOM) has been proposed
in order to remove the "border effect" in conventional Self-Organizing Maps
(SOM). However, SSOM still has limitations in representing a sequence of
events. The concentric spherical Self-Organizing Maps (CSSOM) is proposed in
this report, because it can use an arbitrary number of spheres and that
topology could be applied in analysis of sequential and time series data.
A new method to extend SSOM and to reconstruct the neighbors is
introduced in this paper in order to implement concentric spherical
Self-Organizing Maps. Moreover, for ease of evaluation, the display schemas
and several measurements for the quality of SOMs are also discussed with the
experimental results. The results indicate that the quality of SOM is
improved through using specified CSSOM depending on the characteristics of
the dataset. However, the results for sequence training as currently proposed
needs improvement. Finally, the quality of clustering becomes worse, as the
number of spheres increases and the number of units in each sphere decreases.
Technical Reports <Technical-DOT-Reports-AT-cs-DOT-anu.edu.au>
Last modified: Tue Mar 20 17:43:04 EST 2012