Tighter generalization bounds for matrix completion via factorization into constrained matrices

Ken ichiro Moridomi; Kohei Hatano; Eiji Takimoto

doi:10.1587/transinf.2017EDP7339

Tighter generalization bounds for matrix completion via factorization into constrained matrices

Ken ichiro Moridomi
, Kohei Hatano
, Eiji Takimoto

Research output: Contribution to journal › Article › peer-review

4 Citations (Scopus)

Abstract

We prove generalization error bounds of classes of low-rank matrices with some norm constraints for collaborative filtering tasks. Our bounds are tighter, compared to known bounds using rank or the related quantity only, by taking the additional L₁ and L_∞ constraints into account. Also, we show that our bounds on the Rademacher complexity of the classes are optimal.

Original language	English
Pages (from-to)	1997-2004
Number of pages	8
Journal	IEICE Transactions on Information and Systems
Volume	E101D
Issue number	8
DOIs	https://doi.org/10.1587/transinf.2017EDP7339
Publication status	Published - Aug 2018

All Science Journal Classification (ASJC) codes

Software
Hardware and Architecture
Computer Vision and Pattern Recognition
Electrical and Electronic Engineering
Artificial Intelligence

Access to Document

10.1587/transinf.2017EDP7339

Cite this

@article{8f9ba573924044a4976b156160ac6db1,

title = "Tighter generalization bounds for matrix completion via factorization into constrained matrices",

abstract = "We prove generalization error bounds of classes of low-rank matrices with some norm constraints for collaborative filtering tasks. Our bounds are tighter, compared to known bounds using rank or the related quantity only, by taking the additional L1 and L∞ constraints into account. Also, we show that our bounds on the Rademacher complexity of the classes are optimal.",

author = "Moridomi, \{Ken ichiro\} and Kohei Hatano and Eiji Takimoto",

note = "Publisher Copyright: Copyright {\textcopyright} 2018 The Institute of Electronics.",

year = "2018",

month = aug,

doi = "10.1587/transinf.2017EDP7339",

language = "English",

volume = "E101D",

pages = "1997--2004",

journal = "IEICE Transactions on Information and Systems",

issn = "0916-8532",

publisher = "Institute of Electronics Information Communication Engineers",

number = "8",

}

TY - JOUR

T1 - Tighter generalization bounds for matrix completion via factorization into constrained matrices

AU - Moridomi, Ken ichiro

AU - Hatano, Kohei

AU - Takimoto, Eiji

PY - 2018/8

Y1 - 2018/8

N2 - We prove generalization error bounds of classes of low-rank matrices with some norm constraints for collaborative filtering tasks. Our bounds are tighter, compared to known bounds using rank or the related quantity only, by taking the additional L1 and L∞ constraints into account. Also, we show that our bounds on the Rademacher complexity of the classes are optimal.

AB - We prove generalization error bounds of classes of low-rank matrices with some norm constraints for collaborative filtering tasks. Our bounds are tighter, compared to known bounds using rank or the related quantity only, by taking the additional L1 and L∞ constraints into account. Also, we show that our bounds on the Rademacher complexity of the classes are optimal.

UR - https://www.scopus.com/pages/publications/85052026888

UR - https://www.scopus.com/pages/publications/85052026888#tab=citedBy

U2 - 10.1587/transinf.2017EDP7339

DO - 10.1587/transinf.2017EDP7339

M3 - Article

AN - SCOPUS:85052026888

SN - 0916-8532

VL - E101D

SP - 1997

EP - 2004

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

IS - 8

ER -

Tighter generalization bounds for matrix completion via factorization into constrained matrices

Abstract

All Science Journal Classification (ASJC) codes

Access to Document

Other files and links

Fingerprint

Cite this