On estimation problems based on Type III Longla copulas
Document Type
Lecture
Publication Date
10-31-2024
Abstract
We will explore parameter estimation for a new class of non-symmetric copulas. We provide asymptotic distributions for parameter estimators. We study a simple random sample from the bivariate copula, then extend the work to the stationary Markov chain framework. Asymptotic normality is established for proposed estimators and the MLE. The proof uses the unique properties of the eigenfunctions of the copula’s functional operator and the mixing structure of the Markov chains. A simulation study involving sine, cosine, and FGM copulas is presented. This simulation compares the performance of the proposed estimator with that of other existing estimators.
Relational Format
presentation
Recommended Citation
Qazi, Zamzam, "On estimation problems based on Type III Longla copulas" (2024). Probability & Statistics Seminar. 4.
https://egrove.olemiss.edu/math_statistics/4
