Efficient PINNs via multi-head unimodular regularization of the solutions space

Abstract Non-linear differential equations are a fundamental tool to describe different phenomena in nature. However, we still lack a well-established method to tackle stiff differential equations. Here we present a machine learning framework to facilitate the solution of nonlinear multiscale differ...

Full description

Saved in:
Bibliographic Details
Main Authors: Pedro Tarancón-Álvarez, Pablo Tejerina-Pérez, Raul Jimenez, Pavlos Protopapas
Format: Article
Language:English
Published: Nature Portfolio 2025-08-01
Series:Communications Physics
Online Access:https://doi.org/10.1038/s42005-025-02248-1
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Abstract Non-linear differential equations are a fundamental tool to describe different phenomena in nature. However, we still lack a well-established method to tackle stiff differential equations. Here we present a machine learning framework to facilitate the solution of nonlinear multiscale differential equations and, especially, inverse problems using Physics-Informed Neural Networks (PINNs). This framework is based on what is called multi-head (MH) training, which involves training the network to learn a general space of all solutions for a given set of equations with certain variability, rather than learning a specific solution of the system. This setup is used with a second novel technique that we call Unimodular Regularization (UR) of the latent space of solutions. We show that the multi-head approach, combined with Unimodular Regularization, significantly improves the efficiency of PINNs by facilitating the transfer learning process thereby enabling the finding of solutions for nonlinear, coupled, and multiscale differential equations.
ISSN:2399-3650