Skip to content
DSML.KZ
Publications

Code · Zenodo · June 28, 2026

Physics-Informed Neural Networks for Nonlinear Dispersive Equations

Nurbol Koshkarbayev
Open via DOI

Abstract

Initial reproducibility release for the manuscript: Physics-Informed Neural Networks for Nonlinear Dispersive Equations This release contains the PyTorch implementation used to reproduce the numerical experiments, ablation studies, seed-robustness tables, and Kawahara spectral-error diagnostics reported in the manuscript. Included components: Standard physical-space PINN scripts for KdV, BBM, Kawahara, and Rosenau--KdV. Kawahara Fourier error-spectrum diagnostic. Kawahara phase-shift/spectral-control diagnostic. MIT License and citation metadata.

Model Reduction and Neural NetworksNonlinear Waves and SolitonsNeural Networks and Applicationsphysics-informed neural networksnonlinear dispersive equationsKorteweg-de Vries equation

More by Nurbol Koshkarbayev