Project Grant 2207328

This $300,000 Project Grant from the National Science Foundation's Division of Mathematical Sciences, under the Mathematical and Physical Sciences program (CFDA 47.049), will support the development of next-generation mathematical and algorithmic tools to address two key issues in applying machine learning to statistical modeling of time-evolving complex systems: a shortage of informative training data and the high computational costs of high-dimensional problems. Specifically, the Pennsylvania State University will develop a reduced-order statistical closure model to enhance machine learning-based prediction when time series observations are too short, a dimensionality reduction technique respecting the manifold geometry of dynamical variables, and study the theoretical convergence of a Bayesian machine learning algorithm shown to improve El Niño predictions. The award reflects NSF's mission to strengthen the Nation's scientific enterprise through increasing mathematical and physical sciences knowledge and understanding of major national problems.