Random matrix theory in pattern classification

An application to error estimation

Amin Zollanvari, Edward R. Dougherty

Research output: Chapter in Book/Report/Conference proceedingConference contribution

6 Citations (Scopus)

Abstract

We employed the Random Matrix Theory (RMT) to construct a nearly unbiased estimator of true error rate of linear discriminant analysis (LDA) with ridge estimator of inverse covariance matrix in the multivariate Gaussian model and in small-sample situation. In such a scenario, the performance of the constructed estimator, as measured by Root-Mean-Square (RMS) error, shows consistent improvement over well-known estimators of true error.

Original languageEnglish
Title of host publicationConference Record - Asilomar Conference on Signals, Systems and Computers
PublisherIEEE Computer Society
Pages884-887
Number of pages4
ISBN (Print)9781479923908
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event2013 47th Asilomar Conference on Signals, Systems and Computers - Pacific Grove, CA, United States
Duration: Nov 3 2013Nov 6 2013

Other

Other2013 47th Asilomar Conference on Signals, Systems and Computers
CountryUnited States
CityPacific Grove, CA
Period11/3/1311/6/13

Fingerprint

Error analysis
Pattern recognition
Discriminant analysis
Covariance matrix
Mean square error

Keywords

  • Error Estimation
  • Linear discriminant analysis
  • Linear Discriminant Analysis
  • Small-Sample

ASJC Scopus subject areas

  • Computer Networks and Communications
  • Signal Processing

Cite this

Zollanvari, A., & Dougherty, E. R. (2013). Random matrix theory in pattern classification: An application to error estimation. In Conference Record - Asilomar Conference on Signals, Systems and Computers (pp. 884-887). [6810415] IEEE Computer Society. https://doi.org/10.1109/ACSSC.2013.6810415

Random matrix theory in pattern classification : An application to error estimation. / Zollanvari, Amin; Dougherty, Edward R.

Conference Record - Asilomar Conference on Signals, Systems and Computers. IEEE Computer Society, 2013. p. 884-887 6810415.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Zollanvari, A & Dougherty, ER 2013, Random matrix theory in pattern classification: An application to error estimation. in Conference Record - Asilomar Conference on Signals, Systems and Computers., 6810415, IEEE Computer Society, pp. 884-887, 2013 47th Asilomar Conference on Signals, Systems and Computers, Pacific Grove, CA, United States, 11/3/13. https://doi.org/10.1109/ACSSC.2013.6810415
Zollanvari A, Dougherty ER. Random matrix theory in pattern classification: An application to error estimation. In Conference Record - Asilomar Conference on Signals, Systems and Computers. IEEE Computer Society. 2013. p. 884-887. 6810415 https://doi.org/10.1109/ACSSC.2013.6810415
Zollanvari, Amin ; Dougherty, Edward R. / Random matrix theory in pattern classification : An application to error estimation. Conference Record - Asilomar Conference on Signals, Systems and Computers. IEEE Computer Society, 2013. pp. 884-887
@inproceedings{a4fb4b67178b43cb9ad6cf922b2f6400,
title = "Random matrix theory in pattern classification: An application to error estimation",
abstract = "We employed the Random Matrix Theory (RMT) to construct a nearly unbiased estimator of true error rate of linear discriminant analysis (LDA) with ridge estimator of inverse covariance matrix in the multivariate Gaussian model and in small-sample situation. In such a scenario, the performance of the constructed estimator, as measured by Root-Mean-Square (RMS) error, shows consistent improvement over well-known estimators of true error.",
keywords = "Error Estimation, Linear discriminant analysis, Linear Discriminant Analysis, Small-Sample",
author = "Amin Zollanvari and Dougherty, {Edward R.}",
year = "2013",
doi = "10.1109/ACSSC.2013.6810415",
language = "English",
isbn = "9781479923908",
pages = "884--887",
booktitle = "Conference Record - Asilomar Conference on Signals, Systems and Computers",
publisher = "IEEE Computer Society",
address = "United States",

}

TY - GEN

T1 - Random matrix theory in pattern classification

T2 - An application to error estimation

AU - Zollanvari, Amin

AU - Dougherty, Edward R.

PY - 2013

Y1 - 2013

N2 - We employed the Random Matrix Theory (RMT) to construct a nearly unbiased estimator of true error rate of linear discriminant analysis (LDA) with ridge estimator of inverse covariance matrix in the multivariate Gaussian model and in small-sample situation. In such a scenario, the performance of the constructed estimator, as measured by Root-Mean-Square (RMS) error, shows consistent improvement over well-known estimators of true error.

AB - We employed the Random Matrix Theory (RMT) to construct a nearly unbiased estimator of true error rate of linear discriminant analysis (LDA) with ridge estimator of inverse covariance matrix in the multivariate Gaussian model and in small-sample situation. In such a scenario, the performance of the constructed estimator, as measured by Root-Mean-Square (RMS) error, shows consistent improvement over well-known estimators of true error.

KW - Error Estimation

KW - Linear discriminant analysis

KW - Linear Discriminant Analysis

KW - Small-Sample

UR - http://www.scopus.com/inward/record.url?scp=84901277349&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84901277349&partnerID=8YFLogxK

U2 - 10.1109/ACSSC.2013.6810415

DO - 10.1109/ACSSC.2013.6810415

M3 - Conference contribution

SN - 9781479923908

SP - 884

EP - 887

BT - Conference Record - Asilomar Conference on Signals, Systems and Computers

PB - IEEE Computer Society

ER -