Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

GMLS-Nets: Scientific Machine Learning Methods for Unstructured Data

Conference ·
OSTI ID:1633915

Data fields sampled on irregularly spaced points arise in many applications in the sciences and engineering. For regular grids, Convolutional Neural Networks (CNNs) have been successfully used to gaining benefits from weight sharing and invariances. We generalize CNNs by introducing methods for data on unstructured point clouds based on Generalized Moving Least Squares (GMLS). GMLS is a nonparametric technique for estimating linear bounded functionals from scattered data, and has recently been used in the literature for solving partial differential equations. By parameterizing the GMLS estimator, we obtain learning methods for operators with unstructured stencils. In GMLS-Nets the necessary calculations are local, readily parallelizable, and the estimator is supported by a rigorous approximation theory. We show how the framework may be used for unstructured physical data sets to perform functional regression to identify associated differential operators and to regress quantities of interest. The results suggest the architectures to be an attractive foundation for data-driven model development in scientific machine learning applications

Research Organization:
University of California Santa Barbara
Sponsoring Organization:
USDOE Office of Science (SC), Advanced Scientific Computing Research (ASCR) (SC-21)
DOE Contract Number:
SC0019246
OSTI ID:
1633915
Country of Publication:
United States
Language:
English

Similar Records

GMLS-Nets: A Framework for Learning from Unstructured Data
Conference · Thu Dec 12 23:00:00 EST 2019 · OSTI ID:1633916

GMLS-Nets: A Framework for Learning from Unstructured Data
Technical Report · Thu Sep 05 00:00:00 EDT 2019 · OSTI ID:1763256

Parameterized Pseudo-Differential Operators for Applying CNNs to Unstructured Mesh Data
Software · Sun Oct 10 20:00:00 EDT 2021 · OSTI ID:code-94283

Related Subjects