Assessing the Use of the SOM Technique in Data Mining

D. Marghescu and M.J. Rajanen (Finland)

Keywords

SOM, visualization, evaluation, quality of use

Abstract

The Self-Organizing Maps method is a special type of neural network used in clustering, visualization and abstraction. In this paper, we evaluate to what extent users of the SOM technique are satisfied with this tool in visualizing large amounts of data. The contribution of the paper consists of identifying the factors that influence the quality of use of SOM tools at the three levels considered for analysis: visualization of data, interaction with the tool, and information obtained.

Important Links:



Go Back