Dr. Hongyan Zhou
University of Science and Technology Hebei and University of Florida

University of Florida Astronomy Colloquium - Sept. 27th, 2006

Ensemble Learning for Independent Component Analysis of galactic spectra: Method and Applications

A new method is developed for analyzing galactic spectra based on the recent progress in statistics, Ensemble Learning for Independent Component Analysis (EL-ICA). Important spectral parameters, such as starlight reddening, stellar velocity dispersion, stellar mass, and star formation histories, can be obtained simultaneously. Extensive tests show that the derived parameters are reliable for galaxy spectra with the typical quality of modern spectroscopic survey, such as SDSS. Some applications are presented.