Higher-Order Regularization Learning on Hypergraphs
Adrien Weihs, Andrea Bertozzi, Matthew Thorpe · 2025
Read Higher-Order Regularization Learning on Hypergraphs by Adrien Weihs, Andrea Bertozzi, Matthew Thorpe (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-2510-26533.txt.
Description
Adrien Weihs, Andrea L Bertozzi, and Matthew Thorpe extend a learning method built for hypergraphs. A hypergraph links items in groups, not just in pairs, which suits data where ties come in bunches. Their method, higher-order hypergraph learning, enforces smoothness at several scales using powers of the hypergraph's Laplacian, a matrix that encodes its links.
Weihs and his two co-authors prove that a trimmed version of the method stays consistent, meaning it settles toward the right answer as data grows. They also give explicit rates for that settling when it is used in fully supervised learning. Tests in active learning, where the method picks which items to label, went well. So did tests on data with no underlying geometric shape, which the authors take as a sign of range and robustness.
Catalog record
| CREATOR | Adrien Weihs, Andrea Bertozzi, Matthew Thorpe |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 135 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2510-26533