Metraipolis › archive › dc_51fb487faa22

Optimal approximation of smooth functions by ReLU neural networks — arXiv:2610.22693

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.22693
License (as archived)
CC0-1.0
Archived
2026-10-19T09:21:18.772Z
SHA-256 of the archived text
35884767c49d0b4b69c0672e1af2e3c5fd549e2178fd663d8672fa4844e4c86c
Stable URL of this page
https://metraipolis.com/documento/dc_51fb487faa22

How to cite

Optimal approximation of smooth functions by ReLU neural networks — arXiv:2610.22693 — arXiv (en), https://arxiv.org/abs/2610.22693 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_51fb487faa22 (Metraipolis archive dc_51fb487faa22, sha256 35884767c49d0b4b69c0672e1af2e3c5fd549e2178fd663d8672fa4844e4c86c)
BibTeX
@misc{dc51fb487faa22,
  title = {Optimal approximation of smooth functions by ReLU neural networks — arXiv:2610.22693},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.22693},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_51fb487faa22; sha256 35884767c49d0b4b69c0672e1af2e3c5fd549e2178fd663d8672fa4844e4c86c},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.22693. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

We study uniform approximation of the Hölder ball $\mathcal H_1^s([0,1]^D)$ by fully connected ReLU networks. Except in the one-dimensional Lipschitz case, existing constructions remain separated from the information-theoretic efficiency limit. This article closes the gap for every $s>0$ and $D\ge1$: networks with width $O(N)$, depth $O(L)$, and parameter magnitude $O(1)$ attain uniform approximation error of order $(N^2L^2\log N)^{-s/D}$. No relation between width and depth is required. The construction uses new encoding and decoding mechanisms that exploit higher-order finite-difference bounds for local Taylor coefficients of the target function. As a statistical consequence, empirical risk minimizers over the corresponding clipped network classes attain the minimax regression rate $n^{-2s/(2s+D)}$ without additional logarithmic factors. The paper also gives a detailed exposition of the key approximation techniques, covering both established and new mechanisms.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (stat.ML) ; Machine Learning (cs.LG); Numerical Analysis (math.NA).
Pagina del lavoro: https://arxiv.org/abs/2610.22693 — 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 2026-10-19T09:17:41.696Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).