Partial correlation estimates based on signs

Daniel Vogel, Claudia Köllmann, Roland Fried

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

Abstract

We investigate the Oja sign covariance matrix (Oja SCM) for estimating partial correlations in multivariate data. The Oja SCM estimates directly a multiple of the precision matrix and is based on the concept of Oja signs, which generalise the univariate sign function and obey some form of affine equivariance property. We compare it to the classical MLE as well as to estimates based on two alternative multivariate signs: the marginal sign and the spatial sign.
Original languageEnglish
Title of host publicationProceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering
EditorsJ Heikkonen
PublisherTampere International Center for Signal Processing
Number of pages6
Volume43
Publication statusPublished - 2008

Fingerprint

Covariance matrix

Cite this

Vogel, D., Köllmann, C., & Fried, R. (2008). Partial correlation estimates based on signs. In J. Heikkonen (Ed.), Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering (Vol. 43). Tampere International Center for Signal Processing .

Partial correlation estimates based on signs. / Vogel, Daniel; Köllmann, Claudia; Fried, Roland .

Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering. ed. / J Heikkonen. Vol. 43 Tampere International Center for Signal Processing , 2008.

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

Vogel, D, Köllmann, C & Fried, R 2008, Partial correlation estimates based on signs. in J Heikkonen (ed.), Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering. vol. 43, Tampere International Center for Signal Processing .
Vogel D, Köllmann C, Fried R. Partial correlation estimates based on signs. In Heikkonen J, editor, Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering. Vol. 43. Tampere International Center for Signal Processing . 2008
Vogel, Daniel ; Köllmann, Claudia ; Fried, Roland . / Partial correlation estimates based on signs. Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering. editor / J Heikkonen. Vol. 43 Tampere International Center for Signal Processing , 2008.
@inproceedings{f8dd75bf7dbc4925a9976fba57a31c75,
title = "Partial correlation estimates based on signs",
abstract = "We investigate the Oja sign covariance matrix (Oja SCM) for estimating partial correlations in multivariate data. The Oja SCM estimates directly a multiple of the precision matrix and is based on the concept of Oja signs, which generalise the univariate sign function and obey some form of affine equivariance property. We compare it to the classical MLE as well as to estimates based on two alternative multivariate signs: the marginal sign and the spatial sign.",
author = "Daniel Vogel and Claudia K{\"o}llmann and Roland Fried",
year = "2008",
language = "English",
volume = "43",
editor = "J Heikkonen",
booktitle = "Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering",
publisher = "Tampere International Center for Signal Processing",

}

TY - GEN

T1 - Partial correlation estimates based on signs

AU - Vogel, Daniel

AU - Köllmann, Claudia

AU - Fried, Roland

PY - 2008

Y1 - 2008

N2 - We investigate the Oja sign covariance matrix (Oja SCM) for estimating partial correlations in multivariate data. The Oja SCM estimates directly a multiple of the precision matrix and is based on the concept of Oja signs, which generalise the univariate sign function and obey some form of affine equivariance property. We compare it to the classical MLE as well as to estimates based on two alternative multivariate signs: the marginal sign and the spatial sign.

AB - We investigate the Oja sign covariance matrix (Oja SCM) for estimating partial correlations in multivariate data. The Oja SCM estimates directly a multiple of the precision matrix and is based on the concept of Oja signs, which generalise the univariate sign function and obey some form of affine equivariance property. We compare it to the classical MLE as well as to estimates based on two alternative multivariate signs: the marginal sign and the spatial sign.

M3 - Conference contribution

VL - 43

BT - Proceedings of the 1st Workshop on Information Theoretic Methods in Science and Engineering

A2 - Heikkonen, J

PB - Tampere International Center for Signal Processing

ER -