Multivariate Medial Correlation with applications
Methodology
2020-10-15 v3 Probability
Statistics Theory
Statistics Theory
Abstract
We define a multivariate medial correlation coefficient that extends the probabilistic interpretation and properties of Blomqvist's coefficient, incorporates multivariate marginal dependencies and it preserves a stronger multivariate concordance relation. We determine the maximum and minimum values attainable and illustrate the results in some models. We end with an application on real datasets.
Cite
@article{arxiv.2004.05192,
title = {Multivariate Medial Correlation with applications},
author = {Helena Ferreira and Marta Ferreira},
journal= {arXiv preprint arXiv:2004.05192},
year = {2020}
}