Previous | Session 29 | Next | Author Index | Block Schedule
D. Bazell (Eureka Scientific, Inc.), D. J. Miller (Penn State University)
We analyze a portion of the SDSS photometric catalog consisting of approximately 10,000 objects that have been spectroscopically classifed into stars, galaxies, QSOs, late-type stars and unknown objects in order to investigate the existence and nature of subclasses of the unknown objects. We use a modified mixture modeling approach that makes use of both labeled and unlabeled data and performs class discovery on the data set. This technique discovers potentially new object classes when a new cluster of objects is sufficiently distinct from other clusters in the data set. The new clusters are of possible scientific interest because they represent a structured group of outliers. We identify two well defined subclasses of the unknown object class. One subclass contains 58% unknown objects, 40% stars, and 2% galaxies, QSOs, and late-type stars. The other subclass contains 91% unknown objects, 6% late-type stars, and 3% stars, galaxies, and QSOs. We discuss possible interpretations of these subclasses and identify some cautionary areas for color-based object classification. As a side benefit of this limited study we also find two distinct classes, consisting largely of galaxies, that coincide with the recently discussed bimodal galaxy color distribution.
Previous | Session 29 | Next
Bulletin of the American Astronomical Society, 37 #4
© 2005. The American Astronomical Soceity.