Asymptotically unbiased estimators for the extreme-value index
Estimators of the extreme-value index are based on a set of upper order statistics. When the number of upper-order statistics used in the estimation of the extreme-value index is small, the variance of the estimator will be large. On the other hand, the use of a large number of upper statistics will introduce a big bias. There are several papers concerning how to balance the variance component and the bias component. In this paper, we give an unbiased estimator even if one uses a large number of upper-order statistics.
Year of publication: |
1998
|
---|---|
Authors: | Peng, L. |
Published in: |
Statistics & Probability Letters. - Elsevier, ISSN 0167-7152. - Vol. 38.1998, 2, p. 107-115
|
Publisher: |
Elsevier |
Keywords: | Extreme-value index Hill estimator Pickands' estimator |
Saved in:
Online Resource
Saved in favorites
Similar items by person
-
Using a Bootstrap Method to choose the Sample Fraction in Tail Index Estimation
Danielsson, J., (1997)
-
Fiscal policies, optimal growth, and political elections under different economic systems
Lian, Peng, (1993)
-
Using a bootstrap method to choose the sample fraction in tail index estimation
Daníelsson, Jón, (1997)
- More ...