[Previous] | [Session 4] | [Next]

[4.10] Automated Classification of X-ray Sources in Stellar Clusters

S.M. Hojnacki, J.H. Kastner (Rochester Institute of Technology)

We are adapting and applying multivariate statistics techniques, analogous to those routinely applied to multispectral and hyperspectral terrestrial imagery, in the context of analysis of Chandra X-ray Observatory (CXO) and X-ray Multi-Mirror (XMM) imaging spectroscopy of star formation regions. An automated classification technique is being developed to group pre-main sequence X-ray sources into clusters based on spectral and temporal attributes. The algorithm is being tested on deep ACIS images of the population of approximately 1000 X-ray emitting young stars in the Orion Nebula Cluster (ONC). Preliminary results from clustering on ONC sources demonstrate that an unsupervised method can be used to group X-ray sources into distinct classes, wherein members of a given class display similar spectral features. No a priori knowledge of the nature of each source was used to accomplish the clustering. This research is being funded in part by the Eastman Kodak Company via a grant to the Chester F. Carlson Center for Imaging Science at the Rochester Institute of Technology.

[Previous] | [Session 4] | [Next]

Bulletin of the American Astronomical Society, 35#5
© 2003. The American Astronomical Soceity.