Scholar
Paul J. Atzberger
Google Scholar ID: DhEdDBMAAAAJ
Professor of Mathematics and Mechanical Engineering, University of California Santa Barbara
Citations & Impact
Citations
984
H-index
17
i10-index
21
Publications
20
Co-authors
37
list available
Resume (English only)
Academic Achievements
- Published 'Geometric Neural Operators (GNPs) for Data-Driven Deep Learning in Non-Euclidean Settings', Machine Learning: Science and Technology, 2024
- Preprint 'Transferable Foundation Models for Geometric Tasks on Point Cloud Representations: Geometric Neural Operators', arXiv:2503.04649
- Published 'SDYN-GANs: Adversarial Learning Methods for Multistep Generative Models for General Order Stochastic Dynamics', 2023
- Published 'Variational Autoencoders for Learning Nonlinear Dynamics of Physical Systems', 2020
- Published 'GD-VAEs: Geometric Dynamic Variational Autoencoders for Learning Nonlinear Dynamics and Dimension Reductions', 2022
- Published 'GMLS-Nets: A Framework for Learning from Unstructured Data', arXiv:1909.05371, 2019
- Funded by the National Science Foundation (NSF) under Grants DMS-0635535 and CAREER DMS-0956210
Research Experience
- Leads research on computational methods for hydrodynamics of curved fluid interfaces
- Develops open-source software for fluctuating hydrodynamic simulations
- Works on implicit-solvent coarse-grained models and lipid bilayer membrane hydrodynamics
- Develops GD-VAEs (Geometric Dynamic Variational Autoencoders) for learning nonlinear dynamics and dimension reduction
- Creates Geometric Neural Operators (GNPs) for data-driven deep learning in non-Euclidean settings
- Researches GMLS-Nets framework for learning differential operators from unstructured scattered data
