Metraipolis › archive › dc_07396c4d94da

Sparsely connected rank-inspired neural network — arXiv:2609.30750

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.30750
License (as archived)
CC0-1.0
Archived
2026-10-08T02:58:03.629Z
SHA-256 of the archived text
6cc7f3de33cc21280ba1345835b28bf2711f7bf1fa675fbfd49a9cdbc548c155
Stable URL of this page
https://metraipolis.com/documento/dc_07396c4d94da

How to cite

Sparsely connected rank-inspired neural network — arXiv:2609.30750 — arXiv (en), https://arxiv.org/abs/2609.30750 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_07396c4d94da (Metraipolis archive dc_07396c4d94da, sha256 6cc7f3de33cc21280ba1345835b28bf2711f7bf1fa675fbfd49a9cdbc548c155)
BibTeX
@misc{dc07396c4d94da,
  title = {Sparsely connected rank-inspired neural network — arXiv:2609.30750},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.30750},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_07396c4d94da; sha256 6cc7f3de33cc21280ba1345835b28bf2711f7bf1fa675fbfd49a9cdbc548c155},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.30750. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We propose a Sparsely Connected Rank-Inspired Neural Network (SCRINN) for function approximation and the numerical solution of partial differential equations (PDEs). By introducing a structured sparse connectivity architecture and an orthogonality- and normalization-guided loss function, SCRINN constructs an approximately orthonormal neural basis while promoting progressive rank growth of its evaluation matrix and reducing the number of trainable parameters. A two-stage framework is developed, in which an approximately orthonormal neural basis is first constructed, and the output weights are subsequently determined using least-squares methods. Numerical experiments on function approximation and a variety of steady-state, time-dependent, and nonlinear PDEs demonstrate that SCRINN consistently achieves high approximation accuracy with substantially lower training costs. The code is publicly available on GitHub.

Tema degli esploratori: matematica e informatica teorica (retro-catalogo arXiv: annunci del 2026-09-28 in math.NA).
Categorie arXiv: Numerical Analysis (math.NA).
Pagina del lavoro: https://arxiv.org/abs/2609.30750 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T02:27:28.553Z dalla pagina https://arxiv.org/catchup/math.NA/2026-09-28?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).