Saved in:
Bibliographic Details
Main Authors: Wang, Shihao, Qian, Qipeng, Wang, Jingquan
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.00872
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914303171362816
author Wang, Shihao
Qian, Qipeng
Wang, Jingquan
author_facet Wang, Shihao
Qian, Qipeng
Wang, Jingquan
contents We study solution learning for heat-based equations in self-similar variables (SSV). We develop an SSV training framework compatible with standard neural-operator training. We instantiate this framework on the two-dimensional incompressible Navier-Stokes equations and the one-dimensional viscous Burgers equation, and perform controlled comparisons between models trained in physical coordinates and in the corresponding self-similar coordinates using two simple fully connected architectures (standard multilayer perceptrons and a factorized fully connected network). Across both systems and both architectures, SSV-trained networks consistently deliver substantially more accurate and stable extrapolation beyond the training window and better capture qualitative long-time trends. These results suggest that self-similar coordinates provide a mathematically motivated inductive bias for learning the long-time dynamics of heat-based equations.
format Preprint
id arxiv_https___arxiv_org_abs_2602_00872
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Learning Heat-based Equations in Self-similar variables
Wang, Shihao
Qian, Qipeng
Wang, Jingquan
Machine Learning
Mathematical Physics
We study solution learning for heat-based equations in self-similar variables (SSV). We develop an SSV training framework compatible with standard neural-operator training. We instantiate this framework on the two-dimensional incompressible Navier-Stokes equations and the one-dimensional viscous Burgers equation, and perform controlled comparisons between models trained in physical coordinates and in the corresponding self-similar coordinates using two simple fully connected architectures (standard multilayer perceptrons and a factorized fully connected network). Across both systems and both architectures, SSV-trained networks consistently deliver substantially more accurate and stable extrapolation beyond the training window and better capture qualitative long-time trends. These results suggest that self-similar coordinates provide a mathematically motivated inductive bias for learning the long-time dynamics of heat-based equations.
title Learning Heat-based Equations in Self-similar variables
topic Machine Learning
Mathematical Physics
url https://arxiv.org/abs/2602.00872