ABSTRACT

The Fisher information matrix related to an order statistic and its concomitant used to order a bivariate random sample are obtained in the case of the shape-parameter vector of an iterated Farlie–Gumbel–Morgenstern bivariate distribution. They contain information conveyed by singly or multiply censored bivariate samples drawn from an iterated Farlie–Gumbel–Morgenstern bivariate distribution. Fisher information is computed for the mean of the exponential distribution in the concomitant of an order statistic. Shannon entropy in the order statistics and their concomitants based on the iterated Farlie–Gumbel–Morgenstern bivariate distribution are derived.

摘要

在迭代 Farlie-Gumbel-Morgenstern 双变量分布的形状参数向量的情况下，获得与阶统计量相关的 Fisher 信息矩阵及其用于对双变量随机样本进行排序的伴随物。它们包含由从迭代 Farlie-Gumbel-Morgenstern 双变量分布中抽取的单或多删失双变量样本传达的信息。Fisher 信息是为伴随顺序统计的指数分布的平均值计算的。推导出基于迭代 Farlie-Gumbel-Morgenstern 二元分布的阶次统计及其伴随项中的香农熵。