In this thesis methodologies have been established for applying self-organizing maps in the exploratory analysis of large data sets. The methods have been demonstrated in three different kinds of case studies: continuous-valued dense data (EEG signals), continuous-valued sparse data (indicators of the standard of living of different countries), and discrete-valued data (full-text document collections).
The self-organizing maps illustrate structures in the data in a different manner than, for example, multidimensional scaling, a more traditional multivariate data analysis methodology. The SOM algorithm concentrates on preserving the neighborhood relations in the data instead of trying to preserve the distances between the data items. Comparisons between methods having different goals must eventually be based on their practical applicability. Here the SOM has been shown to provide a viable alternative.