Positive definite estimators of large covariance matrices

Adam J. Rothman

doi:10.1093/biomet/ass025

Positive definite estimators of large covariance matrices

Adam J. Rothman

Statistics (Twin Cities)

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

97 Scopus citations

Abstract

Using convex optimization, we construct a sparse estimator of the covariance matrix that is positive definite and performs well in high-dimensional settings. A lasso-type penalty is used to encourage sparsity and a logarithmic barrier function is used to enforce positive definiteness. Consistency and convergence rate bounds are established as both the number of variables and sample size diverge. An efficient computational algorithm is developed and the merits of the approach are illustrated with simulations and a speech signal classification example.

Original language	English (US)
Pages (from-to)	733-740
Number of pages	8
Journal	Biometrika
Volume	99
Issue number	3
DOIs	https://doi.org/10.1093/biomet/ass025
State	Published - Sep 2012

Bibliographical note

Funding Information:
The author thanks the editor, associate editor, and referees for helpful suggestions, as well as Tiefeng Jiang for a helpful discussion. The author’s research is supported in part by a grant from the U.S. National Science Foundation.

Keywords

Barrier function
Classification
Convex optimization
High-dimensional data
Sparsity

Publisher link

10.1093/biomet/ass025

Find It

Find It at U of M Libraries

Cite this

@article{11c763bb539f46db99819c828f37810f,

title = "Positive definite estimators of large covariance matrices",

abstract = "Using convex optimization, we construct a sparse estimator of the covariance matrix that is positive definite and performs well in high-dimensional settings. A lasso-type penalty is used to encourage sparsity and a logarithmic barrier function is used to enforce positive definiteness. Consistency and convergence rate bounds are established as both the number of variables and sample size diverge. An efficient computational algorithm is developed and the merits of the approach are illustrated with simulations and a speech signal classification example.",

keywords = "Barrier function, Classification, Convex optimization, High-dimensional data, Sparsity",

author = "Rothman, {Adam J.}",

note = "Funding Information: The author thanks the editor, associate editor, and referees for helpful suggestions, as well as Tiefeng Jiang for a helpful discussion. The author{\textquoteright}s research is supported in part by a grant from the U.S. National Science Foundation.",

year = "2012",

month = sep,

doi = "10.1093/biomet/ass025",

language = "English (US)",

volume = "99",

pages = "733--740",

journal = "Biometrika",

issn = "0006-3444",

publisher = "Oxford University Press",

number = "3",

}

TY - JOUR

T1 - Positive definite estimators of large covariance matrices

AU - Rothman, Adam J.

N1 - Funding Information: The author thanks the editor, associate editor, and referees for helpful suggestions, as well as Tiefeng Jiang for a helpful discussion. The author’s research is supported in part by a grant from the U.S. National Science Foundation.

PY - 2012/9

Y1 - 2012/9

N2 - Using convex optimization, we construct a sparse estimator of the covariance matrix that is positive definite and performs well in high-dimensional settings. A lasso-type penalty is used to encourage sparsity and a logarithmic barrier function is used to enforce positive definiteness. Consistency and convergence rate bounds are established as both the number of variables and sample size diverge. An efficient computational algorithm is developed and the merits of the approach are illustrated with simulations and a speech signal classification example.

AB - Using convex optimization, we construct a sparse estimator of the covariance matrix that is positive definite and performs well in high-dimensional settings. A lasso-type penalty is used to encourage sparsity and a logarithmic barrier function is used to enforce positive definiteness. Consistency and convergence rate bounds are established as both the number of variables and sample size diverge. An efficient computational algorithm is developed and the merits of the approach are illustrated with simulations and a speech signal classification example.

KW - Barrier function

KW - Classification

KW - Convex optimization

KW - High-dimensional data

KW - Sparsity

UR - http://www.scopus.com/inward/record.url?scp=84865499369&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84865499369&partnerID=8YFLogxK

U2 - 10.1093/biomet/ass025

DO - 10.1093/biomet/ass025

M3 - Article

AN - SCOPUS:84865499369

SN - 0006-3444

VL - 99

SP - 733

EP - 740

JO - Biometrika

JF - Biometrika

IS - 3

ER -

Positive definite estimators of large covariance matrices

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this