Finite Element Representation Network (FERN) for Operator Learning with a Localized Trainable Basis

Zecheng Zhang, Hao Liu, Guosheng Fu, Hayden Schaeffer, Guang Lin · 2025

Read Finite Element Representation Network (FERN) for Operator Learning with a Localized Trainable Basis by Zecheng Zhang, Hao Liu, Guosheng Fu, Hayden Schaeffer, Guang Lin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2510-26962.txt.

Description

Zecheng Zhang, Hao Liu, Guosheng Fu, Hayden Schaeffer, and Guang Lin propose a learning method for equations that describe change across space and time. Operator learning maps an input function to an output function, often by writing the output as a sum of basis functions, meaning fixed building-block shapes. Global blocks can miss sharp local features such as shocks.

Their network, called FERN, builds local blocks in the style of the finite element method, a classic way to solve such equations piece by piece. With simple activation rules, the blocks form exactly inside the network and can shift to match local features. Tests on seven families of equations show good accuracy with fewer trainable values than deep global methods.

Catalog record

CREATORZecheng Zhang, Hao Liu, Guosheng Fu, Hayden Schaeffer, Guang Lin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE71 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2510-26962

Browse the full library · About the Tower of Babel