Testing Conditional Independence via Quantile Regression Based Partial Copulas

Research output: Contribution to journal › Journal article › Research › peer-review

Standard

Testing Conditional Independence via Quantile Regression Based Partial Copulas. / Petersen, Lasse; Hansen, Niels Richard.

In: Journal of Machine Learning Research, Vol. 22, 2021, p. 1-47.

Research output: Contribution to journal › Journal article › Research › peer-review

Harvard

Petersen, L & Hansen, NR 2021, 'Testing Conditional Independence via Quantile Regression Based Partial Copulas', Journal of Machine Learning Research, vol. 22, pp. 1-47.

APA

Petersen, L., & Hansen, N. R. (2021). Testing Conditional Independence via Quantile Regression Based Partial Copulas. Journal of Machine Learning Research, 22, 1-47.

Vancouver

Petersen L, Hansen NR. Testing Conditional Independence via Quantile Regression Based Partial Copulas. Journal of Machine Learning Research. 2021;22:1-47.

Author

Petersen, Lasse ; Hansen, Niels Richard. / Testing Conditional Independence via Quantile Regression Based Partial Copulas. In: Journal of Machine Learning Research. 2021 ; Vol. 22. pp. 1-47.

Bibtex

@article{e223c47d5d4945c8aa844bc59611ac6e,
title = "Testing Conditional Independence via Quantile Regression Based Partial Copulas",
abstract = "The partial copula provides a method for describing the dependence between two random variables X and Y conditional on a third random vector Z in terms of nonparametric residuals U1 and U2. This paper develops a nonparametric test for conditional independence by combining the partial copula with a quantile regression based method for estimating the nonparametric residuals. We consider a test statistic based on generalized correlation between U1 and U2 and derive its large sample properties under consistency assumptions on the quantile regression procedure. We demonstrate through a simulation study that the resulting test is sound under complicated data generating distributions. Moreover, in the examples considered the test is competitive to other state-of-the-art conditional independence tests in terms of level and power, and it has superior power in cases with conditional variance heterogeneity of X and Y given Z. {\textcopyright} 2021 Lasse Petersen and Niels Richard Hansen.",
author = "Lasse Petersen and Hansen, {Niels Richard}",
year = "2021",
language = "English",
volume = "22",
pages = "1--47",
journal = "Journal of Machine Learning Research",
issn = "1533-7928",
publisher = "MIT Press",

}

RIS

TY - JOUR

T1 - Testing Conditional Independence via Quantile Regression Based Partial Copulas

AU - Petersen, Lasse

AU - Hansen, Niels Richard

PY - 2021

Y1 - 2021

N2 - The partial copula provides a method for describing the dependence between two random variables X and Y conditional on a third random vector Z in terms of nonparametric residuals U1 and U2. This paper develops a nonparametric test for conditional independence by combining the partial copula with a quantile regression based method for estimating the nonparametric residuals. We consider a test statistic based on generalized correlation between U1 and U2 and derive its large sample properties under consistency assumptions on the quantile regression procedure. We demonstrate through a simulation study that the resulting test is sound under complicated data generating distributions. Moreover, in the examples considered the test is competitive to other state-of-the-art conditional independence tests in terms of level and power, and it has superior power in cases with conditional variance heterogeneity of X and Y given Z. © 2021 Lasse Petersen and Niels Richard Hansen.

AB - The partial copula provides a method for describing the dependence between two random variables X and Y conditional on a third random vector Z in terms of nonparametric residuals U1 and U2. This paper develops a nonparametric test for conditional independence by combining the partial copula with a quantile regression based method for estimating the nonparametric residuals. We consider a test statistic based on generalized correlation between U1 and U2 and derive its large sample properties under consistency assumptions on the quantile regression procedure. We demonstrate through a simulation study that the resulting test is sound under complicated data generating distributions. Moreover, in the examples considered the test is competitive to other state-of-the-art conditional independence tests in terms of level and power, and it has superior power in cases with conditional variance heterogeneity of X and Y given Z. © 2021 Lasse Petersen and Niels Richard Hansen.

M3 - Journal article

VL - 22

SP - 1

EP - 47

JO - Journal of Machine Learning Research

JF - Journal of Machine Learning Research

SN - 1533-7928

ER -

ID: 260351057