The LC-prior GP combines POD-reduced coefficients with a physics-corrected prior and RBF-FD data generation to surrogate nonlinear multi-coupled PDEs on irregular 2D domains more efficiently than standard approaches.
Kernel methods are competitive for operator learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
stat.ML 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Gaussian process surrogate with physical law-corrected prior for multi-coupled PDEs defined on irregular geometry
The LC-prior GP combines POD-reduced coefficients with a physics-corrected prior and RBF-FD data generation to surrogate nonlinear multi-coupled PDEs on irregular 2D domains more efficiently than standard approaches.