Scholarlink

<p>This work provides a study of copula parameters estimators based on functions of Markov chains they generate. Copulas of interest are perturbations of the independence copula. We provide asymptotic distributions of maximum likelihood estimators and confidence intervals for parameters of several copula families introduced in Longla (Stat Pap 65:4331–4363, 2024). A set of moment-like estimators is proposed along with a multivariate asymptotic distribution. We investigate the particular case of Markov chains generated by sine copulas, sine-cosine copulas and the extended Farlie-Gumbel-Morgenstern (FGM) copula family. We provide evidence that our extension of the FGM copula does a better job at capturing the dependence in the <i>Bearing Data Set</i>. Some tests of independence of the sample are proposed. A simulation study is provided for the three examples of copula families that we propose from Type I Longla copulas. This simulation study proposes a comparative study of the introduced estimators, the maximum likelihood estimators and the robust estimators of Longla and Peligrad (J Stat Plan Inference 211:90–106, 2024). It showcases advantages of the proposed estimators.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Estimation problems for some perturbations of the independence copula

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月21日

查看全文
  • Martial Longla,
  • Mous-Abou Hamadou

摘要

This work provides a study of copula parameters estimators based on functions of Markov chains they generate. Copulas of interest are perturbations of the independence copula. We provide asymptotic distributions of maximum likelihood estimators and confidence intervals for parameters of several copula families introduced in Longla (Stat Pap 65:4331–4363, 2024). A set of moment-like estimators is proposed along with a multivariate asymptotic distribution. We investigate the particular case of Markov chains generated by sine copulas, sine-cosine copulas and the extended Farlie-Gumbel-Morgenstern (FGM) copula family. We provide evidence that our extension of the FGM copula does a better job at capturing the dependence in the Bearing Data Set. Some tests of independence of the sample are proposed. A simulation study is provided for the three examples of copula families that we propose from Type I Longla copulas. This simulation study proposes a comparative study of the introduced estimators, the maximum likelihood estimators and the robust estimators of Longla and Peligrad (J Stat Plan Inference 211:90–106, 2024). It showcases advantages of the proposed estimators.