Saved in:
Bibliographic Details
Main Authors: Kong, Fanze, Lai, Chen-Chih, Lu, Yubin
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.01854
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909729279705088
author Kong, Fanze
Lai, Chen-Chih
Lu, Yubin
author_facet Kong, Fanze
Lai, Chen-Chih
Lu, Yubin
contents This paper proposes a data-driven learning framework for identifying governing laws of generalized diffusions with non-gradient components. By combining energy dissipation laws with a physically consistent penalty and first-moment evolution, we design a two-stage method to recover the pseudo-potential and rotation in the pointwise orthogonal decomposition of a class of non-gradient drifts in generalized diffusions. Our two-stage method is applied to complex generalized diffusion processes including dissipation-rotation dynamics, rough pseudo-potentials and noisy data. Representative numerical experiments demonstrate the effectiveness of our approach for learning physical laws in non-gradient generalized diffusions.
format Preprint
id arxiv_https___arxiv_org_abs_2508_01854
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Moment Estimate and Variational Approach for Learning Generalized Diffusion with Non-gradient Structures
Kong, Fanze
Lai, Chen-Chih
Lu, Yubin
Computational Physics
Machine Learning
Analysis of PDEs
Adaptation and Self-Organizing Systems
This paper proposes a data-driven learning framework for identifying governing laws of generalized diffusions with non-gradient components. By combining energy dissipation laws with a physically consistent penalty and first-moment evolution, we design a two-stage method to recover the pseudo-potential and rotation in the pointwise orthogonal decomposition of a class of non-gradient drifts in generalized diffusions. Our two-stage method is applied to complex generalized diffusion processes including dissipation-rotation dynamics, rough pseudo-potentials and noisy data. Representative numerical experiments demonstrate the effectiveness of our approach for learning physical laws in non-gradient generalized diffusions.
title Moment Estimate and Variational Approach for Learning Generalized Diffusion with Non-gradient Structures
topic Computational Physics
Machine Learning
Analysis of PDEs
Adaptation and Self-Organizing Systems
url https://arxiv.org/abs/2508.01854