Generalization Error Bounds for Kernel Matrix Completion and Extrapolation

Pere Gimenez-Febrer; Alba Pages-Zamora; Georgios B. Giannakis

doi:10.1109/LSP.2020.2970306

Generalization Error Bounds for Kernel Matrix Completion and Extrapolation

Pere Gimenez-Febrer, Alba Pages-Zamora, Georgios B. Giannakis

Electrical and Computer Engineering

Research output: Contribution to journal › Article

1 Scopus citations

Abstract

Prior information can be incorporated in matrix completion to improve estimation accuracy and extrapolate the missing entries. Reproducing kernel Hilbert spaces provide tools to leverage the said prior information, and derive more reliable algorithms. This paper analyzes the generalization error of such approaches, and presents numerical tests confirming the theoretical results.

Original language	English (US)
Article number	8974415
Pages (from-to)	326-330
Number of pages	5
Journal	IEEE Signal Processing Letters
Volume	27
DOIs	https://doi.org/10.1109/LSP.2020.2970306
State	Published - Jan 1 2020

Keywords

generalization error
kernel regression
Matrix completion
Rademacher complexity

Access

10.1109/LSP.2020.2970306

Fingerprint Dive into the research topics of 'Generalization Error Bounds for Kernel Matrix Completion and Extrapolation'. Together they form a unique fingerprint.

Cite this

Gimenez-Febrer, P., Pages-Zamora, A., & Giannakis, G. B. (2020). Generalization Error Bounds for Kernel Matrix Completion and Extrapolation. IEEE Signal Processing Letters, 27, 326-330. [8974415]. https://doi.org/10.1109/LSP.2020.2970306

@article{138506ee4d17495ba27b2d242e90e359,

title = "Generalization Error Bounds for Kernel Matrix Completion and Extrapolation",

abstract = "Prior information can be incorporated in matrix completion to improve estimation accuracy and extrapolate the missing entries. Reproducing kernel Hilbert spaces provide tools to leverage the said prior information, and derive more reliable algorithms. This paper analyzes the generalization error of such approaches, and presents numerical tests confirming the theoretical results.",

keywords = "generalization error, kernel regression, Matrix completion, Rademacher complexity",

author = "Pere Gimenez-Febrer and Alba Pages-Zamora and Giannakis, {Georgios B.}",

year = "2020",

month = jan,

day = "1",

doi = "10.1109/LSP.2020.2970306",

language = "English (US)",

volume = "27",

pages = "326--330",

journal = "IEEE Signal Processing Letters",

issn = "1070-9908",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

}

TY - JOUR

T1 - Generalization Error Bounds for Kernel Matrix Completion and Extrapolation

AU - Gimenez-Febrer, Pere

AU - Pages-Zamora, Alba

AU - Giannakis, Georgios B.

PY - 2020/1/1

Y1 - 2020/1/1

N2 - Prior information can be incorporated in matrix completion to improve estimation accuracy and extrapolate the missing entries. Reproducing kernel Hilbert spaces provide tools to leverage the said prior information, and derive more reliable algorithms. This paper analyzes the generalization error of such approaches, and presents numerical tests confirming the theoretical results.

AB - Prior information can be incorporated in matrix completion to improve estimation accuracy and extrapolate the missing entries. Reproducing kernel Hilbert spaces provide tools to leverage the said prior information, and derive more reliable algorithms. This paper analyzes the generalization error of such approaches, and presents numerical tests confirming the theoretical results.

KW - generalization error

KW - kernel regression

KW - Matrix completion

KW - Rademacher complexity

UR - http://www.scopus.com/inward/record.url?scp=85082170389&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85082170389&partnerID=8YFLogxK

U2 - 10.1109/LSP.2020.2970306

DO - 10.1109/LSP.2020.2970306

M3 - Article

AN - SCOPUS:85082170389

VL - 27

SP - 326

EP - 330

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

SN - 1070-9908

M1 - 8974415

ER -