Some characterizations based on entropy of order statistics and record values

被引：45

作者：

Baratpour, S. ^{[1
]}

Ahmadi, J. ^{[1
]}

Arghami, N. R. ^{[1
]}

机构：

[1] Ferdowsi Univ Mashhad, Sch Math Sci, Mashhad, Iran

来源：

COMMUNICATIONS IN STATISTICS-THEORY AND METHODS | 2007年 / 36卷 / 1-4期

关键词：

hazard rate function; Laguerre polynomial; minimal repair; Minkowski inequality; orthonormal system; reversed hazard rate function; series (parallel) system;

D O I：

10.1080/03610920600966530

中图分类号：

O21 [概率论与数理统计]; C8 [统计学];

学科分类号：

020208 ; 070103 ; 0714 ;

摘要：

In the literature of information theory, Shannon entropy plays an important role and in the context of reliability theory, order statistics and record values are used for statistical modeling. The aim of this article is characterizing the parent distributions based on Shannon entropy of order statistics and record values. It is shown that the equality of the Shannon information in order statistics or record values can determine uniquely the parent distribution. The exponential distribution is characterized through maximizing Shannon entropy of record values under some constraints. The results are useful in the modeling problems.

引用

页码：47 / 57

页数：11

共 17 条

[1]

[Anonymous], 1965, 1 COURSE FUNCTIONAL

[2]

Arnold B.C., 1998, RECORDS

[3]

Arnold B. C., 1998, A First Course in Order Statistics

[4] Residual entropy and its characterizations in terms of hazard function and mean residual life function [J].