Day 791

Friday.

1411.5039
Introduction to astroML: Machine Leraning for Astrophysics
VenderPlas, Connolly, Ivezic, Gray

Astronomy and astrophysics are witnessing dramatic increases in data volume as detectors, telescopes and computers become more powerful.  During the last decade, sky surveys across the EM spectrum have collected hundreds of terabytes of astronomical data for hundreds of millions of sources.  Over the next decade, the data volume will enter the petabyte domain, and provide accurate measurements for billions of sources.  Astronomy and physics students are not traditionally trained to handle such voluminous and complex data sets.  In this paper, describe astroML; an initiative, based on Python and scikit-learn, to develop a compendium of machine learning tools designed to address the statistical needs of the next generation of students and astronomical surveys.  Introduce astroML and present a number of example applications that are enabled by this package.

1411.5363
The relation between dynamical mass-to-light ratio and color for massive quiescent galaxies out to z~2 and comparison with stellar population synthesis models
van de Sande, Kriek, Franx, Bezanson, van Dokkum

Explore the relation between the dynamical M/L and rest-frame color of massive quiescent galaxies out to z~2.  Use a galaxy sample with measured stellar velocity dispersions in combination with HST and ground-based multi-band photometry.  Sample spans a large range in log (M/L_g) (of 1.6 dex) and log (M/L_K) (of 1.3 dex).  There is a strong, approximately linear correlation between the M/L for different wavebands and rest-frame color.  The root-mean-scatter scatter in log (M/L) residuals implies that it is possible to estimate the M/L with an accuracy of ~0.25 dex from a single rest-frame optical color.  SPS models with a Salpeter stellar IMF cannot simultaneously match M/L vs (g-z) and M/L vs (g-K).  By changing the slope of the IMF, still unable to explain the M/L of the bluest and reddest galaxies.  Find that an IMF with a slope between alpha=2.35 and alpha=1.35 provides the best match.  Also explore a broken IMF with a Salpeter slope at M<Msun and M>4Msun and a slope alpha in the intermediate region.  The data favor a slope of alpha=1.35 over alpha=2.35.  Nonetheless, results show that variations between different SPS models are comparable to the IMF variations.  In the analysis, assume that the variation in M/L and color is driven by differences in age, and that other contributions (e.g., metallicity evolution, DM) are small.  These assumptions may be an important source of uncertainty as galaxies evolve in more complex ways.