Selection criteria for fuzzy unsupervised learning: Applied to market segmentation

The use of unsupervised fuzzy learning methods produces a large number of alternative classifications. This paper presents and analyzes a collection of criteria to select the most suitable of a set of classifications. Segmentations of the clients? portfolio are important in the frame of decision making in marketing, because they allow discovering hidden profiles which would not be detected with other methods and establishing different strategies for each defined segment. In the introduced case, classifications have been obtained via the LAMDA algorithm. The use of these criteria reduces remarkably the search space and gives a decision aid tool for marketing experts.