Estimation problems for some perturbations of the independence copula

Location

Room 321, Hume Hall

Start Date

27-4-2024 10:00 AM

End Date

27-4-2024 11:00 AM

Description

In this work, we consider symmetric copulas with density of the form: ∞ C(u,v) = 1+ i=1 λkφk(u)φk(v) (1) where {φk(x),k ∈ N} is an orthogonal basis of L2(0,1) and the sequence |λk| has a finite number of values or converges to 0. For the finite case, we propose an estimator of the vector λ = (λ1,...,λs) by considering the Markov chains generated by the copula using uniform margins as stationary distribution. In order to determine the confidence interval of parameters, we provide a multivariate central limit theorem for the vector λ. a For s = 2, several examples are considered such as the cosine copula, the sinecosine copula and the Legendre copula. For each of those copulas, we determine the estimators of the parameters and the central limit theorem as well. A simulation study is provided with a comparison to other known estimators such as MLE and that of Longla and Peligrad (2021). [A formatted abstract is also attached.]

Relational Format

conference proceeding

Comments

poster

This document is currently not available here.

Share

COinS
 
Apr 27th, 10:00 AM Apr 27th, 11:00 AM

Estimation problems for some perturbations of the independence copula

Room 321, Hume Hall

In this work, we consider symmetric copulas with density of the form: ∞ C(u,v) = 1+ i=1 λkφk(u)φk(v) (1) where {φk(x),k ∈ N} is an orthogonal basis of L2(0,1) and the sequence |λk| has a finite number of values or converges to 0. For the finite case, we propose an estimator of the vector λ = (λ1,...,λs) by considering the Markov chains generated by the copula using uniform margins as stationary distribution. In order to determine the confidence interval of parameters, we provide a multivariate central limit theorem for the vector λ. a For s = 2, several examples are considered such as the cosine copula, the sinecosine copula and the Legendre copula. For each of those copulas, we determine the estimators of the parameters and the central limit theorem as well. A simulation study is provided with a comparison to other known estimators such as MLE and that of Longla and Peligrad (2021). [A formatted abstract is also attached.]