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

Download:

Current browse context:

math.ST

Change to browse by:

References & Citations

Bookmark

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

Mathematics > Statistics Theory

Title: Online nonparametric regression with Sobolev kernels

Abstract: In this work we investigate the variation of the online kernelized ridge regression algorithm in the setting of $d-$dimensional adversarial nonparametric regression. We derive the regret upper bounds on the classes of Sobolev spaces $W_{p}^{\beta}(\mathcal{X})$, $p\geq 2, \beta>\frac{d}{p}$. The upper bounds are supported by the minimax regret analysis, which reveals that in the cases $\beta> \frac{d}{2}$ or $p=\infty$ these rates are (essentially) optimal. Finally, we compare the performance of the kernelized ridge regression forecaster to the known non-parametric forecasters in terms of the regret rates and their computational complexity as well as to the excess risk rates in the setting of statistical (i.i.d.) nonparametric regression.
Comments: 40 pages, 5 figures, 3 tables (version 2)
Subjects: Statistics Theory (math.ST); Machine Learning (cs.LG); Machine Learning (stat.ML)
Cite as: arXiv:2102.03594 [math.ST]
  (or arXiv:2102.03594v2 [math.ST] for this version)

Submission history

From: Oleksandr Zadorozhnyi [view email]
[v1] Sat, 6 Feb 2021 15:05:14 GMT (1533kb,D)
[v2] Tue, 13 Jul 2021 09:20:16 GMT (1524kb,D)

Link back to: arXiv, form interface, contact.