On singular values of matrices with independent rows

Shahar Mendelson; Alain Pajor

doi:10.3150/bj/1161614945

On singular values of matrices with independent rows

Shahar Mendelson, Alain Pajor

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

33 Citations (Scopus)

Abstract

We present deviation inequalities of random operators of the form N^-1 Σ^N _i=1 X_i ⨂ X_i from the average operator 𝔼(X ⨂ X), where Xi are independent random vectors distributed as X, which is a random vector in ℝⁿ or in ℒ₂. We use these inequalities to estimate the singular values of random matrices with independent rows (without assuming that the entries are independent).

Original language	English
Pages (from-to)	761-773
Number of pages	13
Journal	Bernoulli
Volume	12
Issue number	5
DOIs	https://doi.org/10.3150/bj/1161614945
Publication status	Published - 1 Jan 2006

Access to Document

10.3150/bj/1161614945

Cite this

@article{6478b7596f68410aa1e779334431bfdb,

title = "On singular values of matrices with independent rows",

abstract = "We present deviation inequalities of random operators of the form N-1 ΣN i=1 Xi ⨂ Xi from the average operator 𝔼(X ⨂ X), where Xi are independent random vectors distributed as X, which is a random vector in ℝn or in ℒ2. We use these inequalities to estimate the singular values of random matrices with independent rows (without assuming that the entries are independent).",

keywords = "Random vectors in ℝ, Singular values of integral operators",

author = "Shahar Mendelson and Alain Pajor",

year = "2006",

month = jan,

day = "1",

doi = "10.3150/bj/1161614945",

language = "English",

volume = "12",

pages = "761--773",

journal = "Bernoulli",

issn = "1350-7265",

publisher = "Bernoulli Society for Mathematical Statistics and Probability",

number = "5",

}

TY - JOUR

T1 - On singular values of matrices with independent rows

AU - Mendelson, Shahar

AU - Pajor, Alain

PY - 2006/1/1

Y1 - 2006/1/1

N2 - We present deviation inequalities of random operators of the form N-1 ΣN i=1 Xi ⨂ Xi from the average operator 𝔼(X ⨂ X), where Xi are independent random vectors distributed as X, which is a random vector in ℝn or in ℒ2. We use these inequalities to estimate the singular values of random matrices with independent rows (without assuming that the entries are independent).

AB - We present deviation inequalities of random operators of the form N-1 ΣN i=1 Xi ⨂ Xi from the average operator 𝔼(X ⨂ X), where Xi are independent random vectors distributed as X, which is a random vector in ℝn or in ℒ2. We use these inequalities to estimate the singular values of random matrices with independent rows (without assuming that the entries are independent).

KW - Random vectors in ℝ

KW - Singular values of integral operators

UR - http://www.scopus.com/inward/record.url?scp=44249120165&partnerID=8YFLogxK

U2 - 10.3150/bj/1161614945

DO - 10.3150/bj/1161614945

M3 - Article

SN - 1350-7265

VL - 12

SP - 761

EP - 773

JO - Bernoulli

JF - Bernoulli

IS - 5

ER -

On singular values of matrices with independent rows

Abstract

Access to Document

Other files and links

Fingerprint

Cite this