Morphological Classification of Galaxies by Shapelet Decomposition in the Sloan Digital Sky Survey

We describe application of the `shapelet' linear decomposition of galaxyimages to morphological classification using images of $\sim$ 3000 galaxiesfrom the Sloan Digital Sky Survey. After decomposing the galaxies we perform aprincipal component analysis to reduce the number of dimensions of the shapeletspace to nine. We find that each of these nine principal components containsunique morphological information, and give a description of each principalcomponent's contribution to a galaxy's morphology. We find that galaxies ofdiffering Hubble type separate cleanly in the shapelet space. We apply aGaussian mixture model to the 9-dimensional space spanned by the principalcomponents and use the results as a basis for classification. Using the mixturemodel, we separate galaxies into seven classes and give a description of eachclass's physical and morphological properties. We find that several of themixture model classes correlate well with the traditional Hubble types both intheir morphology and their physical parameters (e.g., color, velocitydispersions, etc.). In addition, we find an additional class of late-typemorphology but with high velocity dispersions and very blue color; most ofthese galaxies exhibit post-starburst activity. This method provides anobjective and quantitative alternative to traditional and subjective visualclassification.Comment: 21 pages, 16 figures, accepted by AJ, minor changes per the referee's comment