Classifying Gamma‐Ray Bursts using Self‐organizing Maps

Open Access
Publisher Website
Google Scholar
Abstract
Using the self-organizing map (SOM) algorithm developed by Kohonen in 1982, we investigate the three gamma-ray burst classes discovered by Mukherjee et al. in 1998 and classify the bursts in the current BATSE catalog. The validity of the classification and the isotropy of the burst classes is examined using well-known statistical methods. We are able to show by this independent unsupervised classifier that two previously observed burst classes are indeed present in BATSE data. We also support the 2000 conclusion of Hakkila et al. that there is no strong statistical signal for the existence of the third class, which might actually be due to an instrumental bias. This result is obtained through our independent analysis method, which is not related to the 2000 analysis of Hakkila et al.

This publication has 19 references indexed in Scilit: