TAIL DEPENDENCE ESTIMATION BASED ON ESTIMATION OF KENDALL DISTRIBUTION FUNCTION VIA RATIONAL BERNSTEIN POLYNOMIALS


Erdoğan M. S. , Susam S. O.

6th International Conference on Advances in Statistics, İzmir, Turkey, 16 - 18 October 2020, pp.27

  • Publication Type: Conference Paper / Summary Text
  • City: İzmir
  • Country: Turkey
  • Page Numbers: pp.27

Abstract

This study introduces the estimation of Kendall distribution function via rational Bernstein polynomials as an alternative to the methods used in the literature. The proposed estimation method here has many advantages such as providing better control of the shape of the curve. The estimation of the lower and upper tail dependence is focused on using the new estimator with rational Bernstein polynomials. A designed Monte Carlo study is used to measure the performance of the new method. The simulation study indicates that the proposed estimator here is preferable according to ASE performance. Moreover, as a special case, the rational estimation method reduces to Bernstein polynomial estimation method where all the weights are equal. The new estimator is performed to estimate the dependence coefficient of real data set as example.