Big Data in astronomy: U-M astroinformatics research group

Cube proposed by: Chris Miller

Unit: LSA: Natural Sciences

About this project: 
Astronomy is a "Big Data" science that benefits from the intersection of computer science and applied mathematics/statistics. Fundamentally, modern astronomy revolves around digital pictures of the Universe. Each pixel can have between 10 and a few thousand attributes. At the image resolution of modern astronomical instrumentation, the entire sky requires a peta-scale database. This explosion of data has focused the attention of data-driven astronomers to better understand, utilize, and collaborate on modern statistical and computational methods to solve otherwise intractable research questions. This emerging field is called Astro-informatics. Our group is designed for researchers who want to connect and scale their computational and statistical algorithms on the wealth of astronomical data. Specifically, astronomy is the domain science, where questions about our Universe can best be answered by combining the observations with advances in imaging analysis, non-parametric statistics, inference through machine learning, and high dimensional hypothesis testing and regression statistics.