Machine Learning at the Flatiron Institute Seminar: Mark Cheung

Title: Applications of Machine Learning in Heliophysics

Abstract: Heliophysics is a systems-approach to studying solar activity and its connection to the solar system, including space weather manifestations such as geomagnetic storms and aurorae. The study of the Sun also informs our understanding of physical processes relevant for other astrophysical objects, including turbulence, magnetic reconnection and particle acceleration. The diversity of heliophysics data (multi-wavelength, multi-modal, multi-messenger, time domain) invites the application of machine learning techniques to a variety of problems. In this talk, we present some use cases (e.g. solar flare prediction, spectropolarimetric inversions, solar coronal magnetic field modeling), techniques (e.g. computer vision, physics-informed neural networks) and ideas for further exploration.