Parametric Estimation by Generalized Moment Method for Extremes

Abstrakt

The aim of the present paper is to make clear the purpose and sense of an introduction of a scalar inference function. We call it the t-score. By means of t-scores, it is possible to introduce new characteristics of the central tendency and variability of probability distributions. The estimates of the new measures can be constructed from the usual estimates of the parameters, but we show a more straightforward and computationally often more simple way: the use of a modification of the generalized moment method. Since the estimates are robust in cases of heavy-tailed distributions, our approach appears to be particularly important in study of extreme events.

Kategorie a štítky

  • Článek v odborném periodiku