We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

math.NA
< prev | next >
new | recent | 1712

Change to browse by:

math

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Mathematics > Numerical Analysis

Title: New error bounds for deep networks using sparse grids

Authors: Hadrien Montanelli, Qiang Du
(Submitted on 23 Dec 2017 (v1), last revised 13 Aug 2018 (this version, v3))
Abstract: We prove a theorem concerning the approximation of multivariate functions by deep ReLU networks. We present new error estimates for which the curse of the dimensionality is lessened by establishing a connection with sparse grids.
Subjects: Numerical Analysis (math.NA)
Cite as: arXiv:1712.08688 [math.NA]
  (or arXiv:1712.08688v3 [math.NA] for this version)

Submission history

From: Hadrien Montanelli [view email]
[v1] Sat, 23 Dec 2017 00:18:32 GMT (22kb)
[v2] Fri, 12 Jan 2018 19:11:01 GMT (42kb)
[v3] Mon, 13 Aug 2018 23:29:08 GMT (43kb)
Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?)

Link back to: arXiv, form interface, contact.