Research

Here are my current research projects:

Operator Learning applied to star forming regions. These active rgions of star formation are characterized by complex physical processes that are difficult to model with traditional numerical methods. Further, the observations lose crucial parameters the are required by traditional models. We are developing neural operators to learn the underlying physics of these regions and to predict future states of the system.

• “Modeling Turbulent and Self-Gravitating Fluids with Fourier Neural Operators.” Poletti, K., Offner, S. & Ward, R. Foundations of Data Science, under review.

Uncertainty Quantification in neural operators. We are interested in developing methods to quantify the uncertainty in neural operators. This is crucial for understanding the reliability of the predictions made by these models and for the genralization of these models.