Star/galaxy classification using Kohonen self-organizing maps

Abstract
The automatic classification of star/galaxy data sets is of growing importance as the size of survey data sets increases. Here we present a classifier based on the Kohonen self-organizing map (SOM). The SOM was trained to classify by type using only a small number of manually classified objects. Galaxy classification accuracies of 98 per cent are achieved for objects with a limiting magnitude of 20.

This publication has 0 references indexed in Scilit: