Measuring Structure from Spiral Galaxy Image
This event occurred in the past
Date and Time
Thursday, November 7, 2013 from 1 p.m. to 2 p.m.
Serra Hall, 212
Modern astronomical observatories have produced image data for hundreds of thousands of galaxies. To help answer many questions about the universe, information about galaxy structure must be inferred from these images. In this talk we present the first automated method that can determine arm-segment structure for spiral galaxies of arbitrary spiral arm configuration. We have run our algorithm on about 30,000 galaxies where human classifications are available from the Galaxy Zoo project and in all available measures, our algorithm agrees with the humans about as well as they agree with each other. This method will enable new astronomical and cosmological studies, such as investigating whether the universe has nonzero net angular momentum and understanding the relationship between galaxy structure and environment.
The talk will also have a Q&A session on Google internships and graduate school applications.
Darren Davis graduated from USD in 2008 with a BA in CS and a 4.0 GPA. He also received the prestigious Goldwater scholarship in 2007. He is now a CS PhD candidate at UC Irvine, where he is developing methods to automatically extract information from images of spiral galaxies. His broader research interests include computer vision, machine learning, and other areas of machine intelligence. He has also worked on several software engineering and machine learning projects at Google through summer internships in 2007, 2008, and 2013.
Maria Cristina Manabat