Nick Winovich
Scientific Machine Learning
Scientific Machine Learning
Sandia National Laboratories, New Mexico
P.O. Box 5800
Albuquerque, NM 87185
Biography
Nick’s research focuses on the intersection of machine learning, probability theory, and partial differential equations with an emphasis on scientific and engineering applications. Prior to joining Sandia, he directed his research toward the construction of operator networks (neural networks designed to approximate operators rather than functions) endowed with predictive uncertainty estimates to help gauge the accuracy of model predictions. His current research is concentrated on the development of reinforcement learning techniques aimed to help guide policies and design strategies involving complex physical systems.
Education
B.A. in Mathematics, University of Notre Dame, May 2012
M.S. in Mathematics, University of Oregon, May 2015
Ph.D. in Mathematics, Purdue University, August 2021
Dissertation
Publications
-
Winovich, N. (2021). Deep Operator Network with Predictive Uncertainty [Conference Presenation]. https://doi.org/10.2172/1890389 Publication ID: 75936
-
Ebeida, M., Abdelkader, A., Amenta, N., Kouri, D.P., Parekh, O., Phillips, C., Winovich, N., & Winovich, N. (2020). Novel Geometric Operations for Linear Programming. https://doi.org/10.2172/1813669 Publication ID: 71776
-
Winovich, N., Rushdi, A.A., Phipps, E., Ray, J., Lin, G., Ebeida, M., & Ebeida, M. (2019). Rigorous Data Fusion for Computationally Expensive Simulations. https://doi.org/10.2172/1560809 Publication ID: 64705