Resumen de Classification TEMP data with self-organizing maps

David Lahoz, Miguel San Miguel Marco

• Given the complexity of the data TEMP, meteorological measurements in height realized with weather balloon, is not habitual its directs use in the models of meteorological prediction. The majority of the models use some type of group of the such data as Analysis of Principal Components, Analysis Cluster and others. In this paper the classification is done with Self- Organizing Maps (SOM)(cf. [1]). Before there has been obtained the number of cluster in the data(cf. [2]). They are calculated the references classes in some data TEMP series measured in weather station of the INM (Instituto Nacional de Meteorología).