There is a small and simple Matlab program that needs enhancements, and a resulting dataset to be used on a SOM clustering and classification. A GUI aplication is to be made. The details are simple: GUI, the present program enhancement, with the program variable conditions at the GUI (import, standardize, calculation results output, SOM options and outputs, classification of clusters through the specified method, identification of original units with the new classes) a SOM applied, all the quality measure provided (graphical e non graphical) then compute a metric to classify the SOM clusters into a computed codebook from the BMU average of the trained network; this codebook will be used to unsupervised classify the computed SOM clusters. A class quality metric should computed, based on the class density of the trained network. The attached file is to be used as an example.
