Graph-Based Meshfree Multi-scale Coarse Space Approximation for Two-Level Schwarz Methods
Yucheng Liu, Tak Shing Au Yeung, Eric T. Chung, Simon See · 2026
Read Graph-Based Meshfree Multi-scale Coarse Space Approximation for Two-Level Schwarz Methods by Yucheng Liu, Tak Shing Au Yeung, Eric T. Chung, Simon See (2026) 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-2603-26149.txt.
Description
Liu and co-authors speed up solvers for flow in porous media with a graph neural network. Two-level Schwarz methods are robust for high-contrast Darcy flow, but their setup needs many costly local eigenvalue solves. Their method replaces that setup with a learned coarse space. The network works on the graph of the system matrix. A weighted distance measure tracks the gap to the target space and bounds the condition number, giving a principled training goal. Tests in 2D and 3D cut setup cost and total solve time.
Catalog record
| CREATOR | Yucheng Liu, Tak Shing Au Yeung, Eric T. Chung, Simon See |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48508 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.26149