Data Science for Imaging the Physical World

Modern imaging techniques in the physical sciences produce large, multi-dimensional data sets. Understanding the scientific information content in these data sets is a significant challenge. Modern data science techniques have the potential to make a transformative impact in how we use these data sets. However, we do not have any current course offerings that bring these two communities together to educate them in these new problems and methods. We are therefore proposing a class that will introduce the large data-based problems in the physical sciences; will introduce basic techniques of data science; and apply these techniques to example problems chosen from current research at Columbia. The class will be directed towards senior undergraduates and masters students in both the physical and data sciences.