The thesis investigates and proves the connection between Pearson’s correlation coefficient and Kendall’s tau for elliptical distributions. Both correlation coefficients are defined and compared. In addition, we define elliptical distributions and show how to present them in a different way. This presentation allows us to prove that the sum of two elliptical distributions with the same non-negative symmetric matrix is also an elliptical distribution. As an application the relation between rho and tau for elliptical distributions is derived.
|