Improved Physics-informed neural networks loss function regularization with a variance-based term
Preprint
- 7 March 2026
- preprint
- Published by Elsevier in SSRN Electronic Journal
Abstract
In machine learning and statistical modeling, the mean square or absolute error is commonly adopted as an error metric, also called a "loss function". Although effective in reducing average error, this approach may fail to address localized outliers, leading to significant inaccuracies in regions with sharp gradients or discontinuities. This issue is particularly evident in physics-informed neural networks (PINNs), where such localized errors are expected and affect the overall solution. To overcome this limitation, we propose a novel loss function that combines the mean and standard deviation of the chosen error metric. By minimizing this combined loss function, the method ensures a more uniform error distribution and reduces the impact of localized high-error regions. The proposed loss function is easy to implement and tested in problems of varying complexity: 1D Poisson equation, unsteady Burger’s equation, 2D linear elastic solid mechanics, and 2D steady Navier Stokes equations. The results demonstrate improved solution quality and lower maximum error compared to the standard mean-based loss, with minimal impact on computational time.Keywords
This publication has 24 references indexed in Scilit:
- Physics-informed neural network-based discovery of hyperelastic constitutive models from extremely scarce dataComputer Methods in Applied Mechanics and Engineering, 2025
- Fourier feature embedded physics-informed neural network-based topology optimization (FF-PINNTO) framework for geometrically nonlinear structuresComputer Methods in Applied Mechanics and Engineering, 2025
- RoePINNs: An integration of advanced CFD solvers with Physics-Informed Neural Networks and application in arterial flow modelingComputer Methods in Applied Mechanics and Engineering, 2025
- Phase-field modeling of fracture with physics-informed deep learningComputer Methods in Applied Mechanics and Engineering, 2024
- Residual-based adaptivity for two-phase flow simulation in porous media using Physics-informed Neural NetworksComputer Methods in Applied Mechanics and Engineering, 2022
- A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanicsComputer Methods in Applied Mechanics and Engineering, 2021
- Neural‐network‐based approximations for solving partial differential equationsCommunications in Numerical Methods in Engineering, 1994
- Robust Estimation of a Location Parameter1992
- Multilayer feedforward networks are universal approximatorsNeural Networks, 1989
- Spectral and finite difference solutions of the Burgers equationComputers & Fluids, 1986