... are

Consider, for instance, creating a model of the whole world...

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... becomes

A similar derivation for the cost function for FVQ is given in [3,7]. For simplicity, we approximate the posterior probability of winners by the best vector only. Without significant additional computational cost all the winners could be taken into account. Individual variances for model vectors could be included as well, and we plan to do this in the future.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

...[6]

We are grateful to the PicSOM project members for kindly providing us with the image data set.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.

... weak

This assumption has been made in [6] prior to modeling each variable category with a different SOM.

.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.