Probit transformation for nonparametric kernel estimation of the copula density. Geenens, G., Charpentier, A., & Paindaveine, D. Bernoulli, 23(3):1848–1873, August, 2017. Publisher: Bernoulli Society for Mathematical Statistics and Probability![Probit transformation for nonparametric kernel estimation of the copula density [link]](https://bibbase.org/img/filetypes/link.svg "link")
Paper doi abstract bibtex Copula modeling has become ubiquitous in modern statistics. Here, the problem of nonparametrically estimating a copula density is addressed. Arguably the most popular nonparametric density estimator, the kernel estimator is not suitable for the unit-square-supported copula densities, mainly because it is heavily affected by boundary bias issues. In addition, most common copulas admit unbounded densities, and kernel methods are not consistent in that case. In this paper, a kernel-type copula density estimator is proposed. It is based on the idea of transforming the uniform marginals of the copula density into normal distributions via the probit function, estimating the density in the transformed domain, which can be accomplished without boundary problems, and obtaining an estimate of the copula density through back-transformation. Although natural, a raw application of this procedure was, however, seen not to perform very well in the earlier literature. Here, it is shown that, if combined with local likelihood density estimation methods, the idea yields very good and easy to implement estimators, fixing boundary issues in a natural way and able to cope with unbounded copula densities. The asymptotic properties of the suggested estimators are derived, and a practical way of selecting the crucially important smoothing parameters is devised. Finally, extensive simulation studies and a real data analysis evidence their excellent performance compared to their main competitors.
@article{geenens_probit_2017,
title = {Probit transformation for nonparametric kernel estimation of the copula density},
volume = {23},
issn = {1350-7265},
url = {https://projecteuclid.org/journals/bernoulli/volume-23/issue-3/Probit-transformation-for-nonparametric-kernel-estimation-of-the-copula-density/10.3150/15-BEJ798.full},
doi = {10.3150/15-BEJ798},
abstract = {Copula modeling has become ubiquitous in modern statistics. Here, the problem of nonparametrically estimating a copula density is addressed. Arguably the most popular nonparametric density estimator, the kernel estimator is not suitable for the unit-square-supported copula densities, mainly because it is heavily affected by boundary bias issues. In addition, most common copulas admit unbounded densities, and kernel methods are not consistent in that case. In this paper, a kernel-type copula density estimator is proposed. It is based on the idea of transforming the uniform marginals of the copula density into normal distributions via the probit function, estimating the density in the transformed domain, which can be accomplished without boundary problems, and obtaining an estimate of the copula density through back-transformation. Although natural, a raw application of this procedure was, however, seen not to perform very well in the earlier literature. Here, it is shown that, if combined with local likelihood density estimation methods, the idea yields very good and easy to implement estimators, fixing boundary issues in a natural way and able to cope with unbounded copula densities. The asymptotic properties of the suggested estimators are derived, and a practical way of selecting the crucially important smoothing parameters is devised. Finally, extensive simulation studies and a real data analysis evidence their excellent performance compared to their main competitors.},
number = {3},
urldate = {2023-05-21},
journal = {Bernoulli},
author = {Geenens, Gery and Charpentier, Arthur and Paindaveine, Davy},
month = aug,
year = {2017},
note = {Publisher: Bernoulli Society for Mathematical Statistics and Probability},
keywords = {Boundary bias, copula density, local likelihood density estimation, transformation kernel density estimator, unbounded density},
pages = {1848--1873},
file = {Full Text PDF:/Users/soumikp/Zotero/storage/YXSGY535/Geenens et al. - 2017 - Probit transformation for nonparametric kernel est.pdf:application/pdf;Full Text PDF:/Users/soumikp/Zotero/storage/NHVQ93HS/Geenens et al. - 2017 - Probit transformation for nonparametric kernel est.pdf:application/pdf;suppdf_1.pdf:/Users/soumikp/Zotero/storage/BZMVDMBE/suppdf_1.pdf:application/pdf},
}
Downloads: 0
{"_id":"neBWLJfPiv7kEhFQr","bibbaseid":"geenens-charpentier-paindaveine-probittransformationfornonparametrickernelestimationofthecopuladensity-2017","author_short":["Geenens, G.","Charpentier, A.","Paindaveine, D."],"bibdata":{"bibtype":"article","type":"article","title":"Probit transformation for nonparametric kernel estimation of the copula density","volume":"23","issn":"1350-7265","url":"https://projecteuclid.org/journals/bernoulli/volume-23/issue-3/Probit-transformation-for-nonparametric-kernel-estimation-of-the-copula-density/10.3150/15-BEJ798.full","doi":"10.3150/15-BEJ798","abstract":"Copula modeling has become ubiquitous in modern statistics. Here, the problem of nonparametrically estimating a copula density is addressed. Arguably the most popular nonparametric density estimator, the kernel estimator is not suitable for the unit-square-supported copula densities, mainly because it is heavily affected by boundary bias issues. In addition, most common copulas admit unbounded densities, and kernel methods are not consistent in that case. In this paper, a kernel-type copula density estimator is proposed. It is based on the idea of transforming the uniform marginals of the copula density into normal distributions via the probit function, estimating the density in the transformed domain, which can be accomplished without boundary problems, and obtaining an estimate of the copula density through back-transformation. Although natural, a raw application of this procedure was, however, seen not to perform very well in the earlier literature. Here, it is shown that, if combined with local likelihood density estimation methods, the idea yields very good and easy to implement estimators, fixing boundary issues in a natural way and able to cope with unbounded copula densities. The asymptotic properties of the suggested estimators are derived, and a practical way of selecting the crucially important smoothing parameters is devised. Finally, extensive simulation studies and a real data analysis evidence their excellent performance compared to their main competitors.","number":"3","urldate":"2023-05-21","journal":"Bernoulli","author":[{"propositions":[],"lastnames":["Geenens"],"firstnames":["Gery"],"suffixes":[]},{"propositions":[],"lastnames":["Charpentier"],"firstnames":["Arthur"],"suffixes":[]},{"propositions":[],"lastnames":["Paindaveine"],"firstnames":["Davy"],"suffixes":[]}],"month":"August","year":"2017","note":"Publisher: Bernoulli Society for Mathematical Statistics and Probability","keywords":"Boundary bias, copula density, local likelihood density estimation, transformation kernel density estimator, unbounded density","pages":"1848–1873","file":"Full Text PDF:/Users/soumikp/Zotero/storage/YXSGY535/Geenens et al. - 2017 - Probit transformation for nonparametric kernel est.pdf:application/pdf;Full Text PDF:/Users/soumikp/Zotero/storage/NHVQ93HS/Geenens et al. - 2017 - Probit transformation for nonparametric kernel est.pdf:application/pdf;suppdf_1.pdf:/Users/soumikp/Zotero/storage/BZMVDMBE/suppdf_1.pdf:application/pdf","bibtex":"@article{geenens_probit_2017,\n\ttitle = {Probit transformation for nonparametric kernel estimation of the copula density},\n\tvolume = {23},\n\tissn = {1350-7265},\n\turl = {https://projecteuclid.org/journals/bernoulli/volume-23/issue-3/Probit-transformation-for-nonparametric-kernel-estimation-of-the-copula-density/10.3150/15-BEJ798.full},\n\tdoi = {10.3150/15-BEJ798},\n\tabstract = {Copula modeling has become ubiquitous in modern statistics. Here, the problem of nonparametrically estimating a copula density is addressed. Arguably the most popular nonparametric density estimator, the kernel estimator is not suitable for the unit-square-supported copula densities, mainly because it is heavily affected by boundary bias issues. In addition, most common copulas admit unbounded densities, and kernel methods are not consistent in that case. In this paper, a kernel-type copula density estimator is proposed. It is based on the idea of transforming the uniform marginals of the copula density into normal distributions via the probit function, estimating the density in the transformed domain, which can be accomplished without boundary problems, and obtaining an estimate of the copula density through back-transformation. Although natural, a raw application of this procedure was, however, seen not to perform very well in the earlier literature. Here, it is shown that, if combined with local likelihood density estimation methods, the idea yields very good and easy to implement estimators, fixing boundary issues in a natural way and able to cope with unbounded copula densities. The asymptotic properties of the suggested estimators are derived, and a practical way of selecting the crucially important smoothing parameters is devised. Finally, extensive simulation studies and a real data analysis evidence their excellent performance compared to their main competitors.},\n\tnumber = {3},\n\turldate = {2023-05-21},\n\tjournal = {Bernoulli},\n\tauthor = {Geenens, Gery and Charpentier, Arthur and Paindaveine, Davy},\n\tmonth = aug,\n\tyear = {2017},\n\tnote = {Publisher: Bernoulli Society for Mathematical Statistics and Probability},\n\tkeywords = {Boundary bias, copula density, local likelihood density estimation, transformation kernel density estimator, unbounded density},\n\tpages = {1848--1873},\n\tfile = {Full Text PDF:/Users/soumikp/Zotero/storage/YXSGY535/Geenens et al. - 2017 - Probit transformation for nonparametric kernel est.pdf:application/pdf;Full Text PDF:/Users/soumikp/Zotero/storage/NHVQ93HS/Geenens et al. - 2017 - Probit transformation for nonparametric kernel est.pdf:application/pdf;suppdf_1.pdf:/Users/soumikp/Zotero/storage/BZMVDMBE/suppdf_1.pdf:application/pdf},\n}\n\n","author_short":["Geenens, G.","Charpentier, A.","Paindaveine, D."],"key":"geenens_probit_2017","id":"geenens_probit_2017","bibbaseid":"geenens-charpentier-paindaveine-probittransformationfornonparametrickernelestimationofthecopuladensity-2017","role":"author","urls":{"Paper":"https://projecteuclid.org/journals/bernoulli/volume-23/issue-3/Probit-transformation-for-nonparametric-kernel-estimation-of-the-copula-density/10.3150/15-BEJ798.full"},"keyword":["Boundary bias","copula density","local likelihood density estimation","transformation kernel density estimator","unbounded density"],"metadata":{"authorlinks":{}},"html":""},"bibtype":"article","biburl":"https://bibbase.org/f/Ceciz2iNjTZgQNtDc/mypubs_mar_2024.bib","dataSources":["m8Y57GfgnRrMKZTQS","epk5yKhDyD37NAsSC"],"keywords":["boundary bias","copula density","local likelihood density estimation","transformation kernel density estimator","unbounded density"],"search_terms":["probit","transformation","nonparametric","kernel","estimation","copula","density","geenens","charpentier","paindaveine"],"title":"Probit transformation for nonparametric kernel estimation of the copula density","year":2017}