next up previous
Next: Conclusions Up: Example data sets Previous: Example data sets

Training and linquistic conclusions

Our goal was to locate and identify groups of individuals having similar opinions to the statements. For each statement category, G, T, S, I, A and N, a separate Self-Organizing Map was trained with the 20 fuzzified variables created from each category (except 16 for N). Several distinct neuron groups could easily be located from these SOMs. There were two obvious groups in most maps: those neurons representing individuals having either positive or negative opinions to all of the five (or four) statements. Additional groups were added as needed (see 'Negative attitude to school work, otherwise positive' in figure 4). There were a total of 16 groups located from the 6 categories.

   figure82
Figure 4: Three located groups from category G.

The final Self-Organizing Map was trained with the grouping information received from the other SOMs using several preprocessing methods. As a result, linquistic conclusions similar to the ones shown in figure 5 a) were obtained. For example, the upper right corner of figure 5 a) represents a group of 21 students who have a positive attitude to categories G, T, S and I. Figure 5 b) presents the classification results obtained using different membership functions from the 4th layer of a TS-SOM [4, 5] (256 neurons) compared to binary memberships with and without a loss group.

   figure95
Figure 5: a) Linquistic conclusions. b) Classification estimates.



Anssi Lensu
Tue Nov 3 12:18:16 EET 1998