A holomorphic Kolmogorov-Arnold network framework for solving elliptic problems on arbitrary 2D domains

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Calafà, Matteo, Andriollo, Tito, Engsig-Karup, Allan P., Jeong, Cheol-Ho
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918107758460928
author Calafà, Matteo
Andriollo, Tito
Engsig-Karup, Allan P.
Jeong, Cheol-Ho
author_facet Calafà, Matteo
Andriollo, Tito
Engsig-Karup, Allan P.
Jeong, Cheol-Ho
contents Physics-informed holomorphic neural networks (PIHNNs) have recently emerged as efficient surrogate models for solving differential problems. By embedding the underlying problem structure into the network, PIHNNs require training only to satisfy boundary conditions, often resulting in significantly improved accuracy and computational efficiency compared to traditional physics-informed neural networks (PINNs). In this work, we improve and extend the application of PIHNNs to two-dimensional problems. First, we introduce a novel holomorphic network architecture based on the Kolmogorov-Arnold representation (PIHKAN), which achieves higher accuracy with reduced model complexity. Second, we develop mathematical extensions that broaden the applicability of PIHNNs to a wider class of elliptic partial differential equations, including the Helmholtz equation. Finally, we propose a new method based on Laurent series theory that enables the application of holomorphic networks to multiply-connected plane domains, thereby removing the previous limitation to simply-connected geometries.
format Preprint
id arxiv_https___arxiv_org_abs_2507_22678
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A holomorphic Kolmogorov-Arnold network framework for solving elliptic problems on arbitrary 2D domains
Calafà, Matteo
Andriollo, Tito
Engsig-Karup, Allan P.
Jeong, Cheol-Ho
Computational Engineering, Finance, and Science
Physics-informed holomorphic neural networks (PIHNNs) have recently emerged as efficient surrogate models for solving differential problems. By embedding the underlying problem structure into the network, PIHNNs require training only to satisfy boundary conditions, often resulting in significantly improved accuracy and computational efficiency compared to traditional physics-informed neural networks (PINNs). In this work, we improve and extend the application of PIHNNs to two-dimensional problems. First, we introduce a novel holomorphic network architecture based on the Kolmogorov-Arnold representation (PIHKAN), which achieves higher accuracy with reduced model complexity. Second, we develop mathematical extensions that broaden the applicability of PIHNNs to a wider class of elliptic partial differential equations, including the Helmholtz equation. Finally, we propose a new method based on Laurent series theory that enables the application of holomorphic networks to multiply-connected plane domains, thereby removing the previous limitation to simply-connected geometries.
title A holomorphic Kolmogorov-Arnold network framework for solving elliptic problems on arbitrary 2D domains
topic Computational Engineering, Finance, and Science
url https://arxiv.org/abs/2507.22678