ASSESSMENT OF CLUSTERANALYSIS AND SELF-ORGANIZING MAPS

Market segmentation represents a central problem of preparing marketin g activities. The methodical approach of this problem is supported by clustering methods. Available data are used to detect common grounds r egarding their quality structures. Therefore statistics provides vario us methods for cluster analysis. Self-organizing maps are another poss ibility to form classes. They are a special approach of the artificial neural networks. The statistical methods and these methods, which are based on organic processes of our brain, offer different solutions al though the starting conditions are the same. Often decisions about inv estigations are based on such solutions. Therefore the results of clus tering are very important to reveal systematic information about the s ize of classes and their structure. Methodical notes are needed for th e use of any clustering method. This paper offers a simplified way to select the best result for clustering.