I conducted this work as an undergraduate research assistant at the Laboratory for Atmospheric and Space Physics (LASP) during my undergraduate at the University of Colorado Boulder. We presented this work as a poster presentation at the NeurIPS 3rd Machine Learning for the Physical Sciences workshop in 2020, as well as at the Machine Learning in Heliophysics Conference hosted in March of 2022.
In this work, I implemented K-means and convolutional neural networks in Python with PyTorch to segment polar coronal holes in images of the Sun. I also applied data assimilation methods (3D-Var) to model the fluid dynamics of stellar atmospheres.
Read the original NeurIPS paper on ML4PhysicalSciences workshop page