PROPERTIES OF THE MAXIMUM LIKELIHOOD ESTIMATES OF THE EXPONENT OF PARETO DISTRIBUTION

This paper investigates the statistical properties of maximum likelihood estimation index of the Pareto distribution. In recent years, power distribution laws such as Pareto distribution attract the attention of researchers in various fields of science and technology, from economics and linguistics...

Full description

Saved in:
Bibliographic Details
Main Authors: I. S. Pulkin, A. Tatarintsev
Format: Article
Language:Russian
Published: MIREA - Russian Technological University 2018-12-01
Series:Российский технологический журнал
Subjects:
Online Access:https://www.rtj-mirea.ru/jour/article/view/135
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This paper investigates the statistical properties of maximum likelihood estimation index of the Pareto distribution. In recent years, power distribution laws such as Pareto distribution attract the attention of researchers in various fields of science and technology, from economics and linguistics to Internet traffic analysis. Therefore, the problem of determining the exponent of the power law for a given sample is of exceptional practical importance. It is analytically proved that this estimate is biased, although valid. A formula that eliminates the bias is proposed. Besides, a formula for the variance of the unbiased estimate is analytically derived. In addition, the problem of finding the distribution function and probability density of this estimate as a random variable is set and analytically solved. Next, a formula for mathematical expectation and dispersion based on previously determined probability density is found. The obtained results can be used in various fields of human activity, for example, to predict the intensity of natural and man-made disasters.
ISSN:2782-3210
2500-316X