Orthogonal Discrepancy Kernels for Learning with Partial Physics
Swapnil Manna, Timothy J. Rogers, Lawrence Bull
stat.ML, cs.LG, eess.SP
Submitted: 2026-06-19
License: http://creativecommons.org/licenses/by-nc-nd/4.0/
The gist: We introduce a semi-parametric framework for nonlinear system identification, which decouples discrepancy functions from physics-based components.
Terminology
Abstract
We introduce a semi-parametric framework for nonlinear system identification, which decouples discrepancy functions from physics-based components. Orthogonal Gaussian process regression balances sparse parameter selection (the white box) with discrepancy learning (the black box) to produce interpretable models from incomplete physics.
Related papers
- Behavior of prediction performance metrics with rare events
- Optimal Estimation of Generic Dynamics by Path-Dependent Neural Jump ODEs
- A Posterior-Dynamics Framework for Imaging Inverse Problems with Pretrained Diffusion Priors
- Scalable Posterior Uncertainty for Flexible Density-Based Clustering
- A Transport Geometry of Benign Overfitting
- Efficient quadratic entropy with distance sketches