Advancing Astrophysics with Machine Learning ~ Dr. Jessica Hislop ~ University of Sussex

Abstract: There is a problem in astrophysics and indeed in many disciplines such that we have a lot of data, but we’re not quite sure what to do with it. Looking for trends and correlations in data is the basis of science, but sometimes human intuition is actually a hinderance in these tasks, and this is where machines come in. Machines have the power to find things in the data humans may miss and they work tirelessly (no coffee breaks or sleep!) to then apply this to millions of objects. In this talk, I will present two ways in which I have been recently using machine learning - one is looking at galaxy environments and the other is constraining cosmological parameters. Without delving into the fundamentals of the physics of these tasks, I will present to you the basic problems and then demonstrate the ways in which machine learning can be used. I will also demonstrate to you the power of machine learning in many different tasks, both in and outside of science.