I am a current student in the Baskin School of Engineering Master of Computer Science program at University of California Santa Cruz. My research interests are in machine learning: applications, implementations, bias, and interpretability. In particular, I am interested in exploring how machine learning can be employed within the domains of environmental science, and astronomy.
I see machine learning as a powerful tool which can aid scientific inquiry in a variety of domains. This is especially true of highly interpretable models, which help us improve our understanding of learned relationships in ways that 'black box' models can't.
This project seeks to estimate atmospheric Carbonyl Sulfide using GA2M, a highly interpretable generalized additive model with pairwise interactions.
Carbonyl Sulfide is a long lived gas that is destroyed by the same enzymes that destroy CO2 during photosynthesis. By examining OCS we can get an estimate of how much photosynthesis is occurring over a large area -- something that is difficult to do using CO2. In turn, these estimates can help refine climate change models and forecasts. This project seeks to examine a discrepancy in the global OCS inventory, which currently has a large missing source.