A fuzzy self-organizing map neural network for market segmentation of credit card
To date, the proposed clustering analysis methods are tremendous. In most of the methods, however, human-made determinations, such as the number of clustering groups, should be decided previously. Not only is the result affected by the subjective viewpoint of the decision-maker, but also the cluster...
Saved in:
Published in | Smc 2000 conference proceedings. 2000 ieee international conference on systems, man and cybernetics. 'cybernetics evolving to systems, humans, organizations, and their complex interactions' (cat. no.0 Vol. 5; pp. 3617 - 3622 vol.5 |
---|---|
Main Authors | , , |
Format | Conference Proceeding |
Language | English |
Published |
IEEE
2000
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | To date, the proposed clustering analysis methods are tremendous. In most of the methods, however, human-made determinations, such as the number of clustering groups, should be decided previously. Not only is the result affected by the subjective viewpoint of the decision-maker, but also the clustering efficiency is not good enough. To overcome these drawbacks, this research attempts to combine fuzzy set theory with the unsupervised learning network model to create an unsupervised fuzzy self-organizing map (FSOM) model. This model integrates an artificial neural network with fuzzy set theory to take respective advantages of the learning function and the capability of handling uncertainty problems in human recognition processes. Generally, the fuzzy clustering analysis model developed in this research can completely explain the results from experiments. In addition, this model seems more useful and practical than other clustering methods. The integration of FSOM and backpropagation neural networks to establish an intelligent decision support system can improve the problem of being unable to quickly analyze new customer information and effectively respond by making a suggestion to the decision-maker. |
---|---|
ISBN: | 9780780365834 0780365836 |
ISSN: | 1062-922X 2577-1655 |
DOI: | 10.1109/ICSMC.2000.886571 |