On multivariate Gaussian copulas

Ivan Zezula

doi:10.1016/j.jspi.2009.05.039

On multivariate Gaussian copulas

Ivan Zezula^*

^*Corresponding author for this work

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

46 Citations (Scopus)

Abstract

Gaussian copulas are handy tool in many applications. However, when dimension of data is large, there are too many parameters to estimate. Use of special variance structure can facilitate the task. In many cases, especially when different data types are used. Pearson correlation is not a suitable measure of dependence. We study the properties of Kendall and Spearman correlation coefficients-which have better properties and are invariant under monotone transformations-used at the place of Pearson coefficients. Spearman correlation coefficient appears to be more suitable for use in such complex applications. (C) 2009 Elsevier B.V. All rights reserved.

Original language	English
Pages (from-to)	3942-3946
Number of pages	5
Journal	Journal of Statistical Planning and Inference
Volume	139
Issue number	11
DOIs	https://doi.org/10.1016/j.jspi.2009.05.039
Publication status	Published - 1-Nov-2009
Event	8th Tartu Conference on Multivariate Statistics/6th Conference on Multivariate Distributions with Fixed Marginals - Tartu, Estonia Duration: 26-Jun-2007 → 29-Jun-2007

Keywords

Copulas
Correlation
Kendall
Spearman

Access to Document

10.1016/j.jspi.2009.05.039

Handle.net

Persistent link

Embargoed Document

https://www.scopus.com/record/display.uri?eid=2-s2.0-77955034068&origin=inward&txGid=cbf80d53d09f3538b5e522056c164ca5
Final publisher's version, 333 KB
Request copy

Cite this

@article{83c5f4b91c234b32a3011ad484d50684,

title = "On multivariate Gaussian copulas",

abstract = "Gaussian copulas are handy tool in many applications. However, when dimension of data is large, there are too many parameters to estimate. Use of special variance structure can facilitate the task. In many cases, especially when different data types are used. Pearson correlation is not a suitable measure of dependence. We study the properties of Kendall and Spearman correlation coefficients-which have better properties and are invariant under monotone transformations-used at the place of Pearson coefficients. Spearman correlation coefficient appears to be more suitable for use in such complex applications. (C) 2009 Elsevier B.V. All rights reserved.",

keywords = "Copulas, Correlation, Kendall, Spearman",

author = "Ivan Zezula",

year = "2009",

month = nov,

day = "1",

doi = "10.1016/j.jspi.2009.05.039",

language = "English",

volume = "139",

pages = "3942--3946",

journal = "Journal of Statistical Planning and Inference",

issn = "0378-3758",

publisher = "ELSEVIER SCIENCE BV",

number = "11",

note = "8th Tartu Conference on Multivariate Statistics/6th Conference on Multivariate Distributions with Fixed Marginals ; Conference date: 26-06-2007 Through 29-06-2007",

}

TY - JOUR

T1 - On multivariate Gaussian copulas

AU - Zezula, Ivan

PY - 2009/11/1

Y1 - 2009/11/1

N2 - Gaussian copulas are handy tool in many applications. However, when dimension of data is large, there are too many parameters to estimate. Use of special variance structure can facilitate the task. In many cases, especially when different data types are used. Pearson correlation is not a suitable measure of dependence. We study the properties of Kendall and Spearman correlation coefficients-which have better properties and are invariant under monotone transformations-used at the place of Pearson coefficients. Spearman correlation coefficient appears to be more suitable for use in such complex applications. (C) 2009 Elsevier B.V. All rights reserved.

AB - Gaussian copulas are handy tool in many applications. However, when dimension of data is large, there are too many parameters to estimate. Use of special variance structure can facilitate the task. In many cases, especially when different data types are used. Pearson correlation is not a suitable measure of dependence. We study the properties of Kendall and Spearman correlation coefficients-which have better properties and are invariant under monotone transformations-used at the place of Pearson coefficients. Spearman correlation coefficient appears to be more suitable for use in such complex applications. (C) 2009 Elsevier B.V. All rights reserved.

KW - Copulas

KW - Correlation

KW - Kendall

KW - Spearman

U2 - 10.1016/j.jspi.2009.05.039

DO - 10.1016/j.jspi.2009.05.039

M3 - Article

AN - SCOPUS:68049114488

SN - 0378-3758

VL - 139

SP - 3942

EP - 3946

JO - Journal of Statistical Planning and Inference

JF - Journal of Statistical Planning and Inference

IS - 11

T2 - 8th Tartu Conference on Multivariate Statistics/6th Conference on Multivariate Distributions with Fixed Marginals

Y2 - 26 June 2007 through 29 June 2007

ER -

On multivariate Gaussian copulas

Abstract

Keywords

Access to Document

Handle.net

Embargoed Document

Fingerprint

Cite this