Statistical Analysis for Geometric Distribution Based on Records

摘要

In this paper, based on records from the geometric distribution, non-Bayesian and Bayesian estimators for the unknown parameter are obtained. The empirical Bayes estimate is derived using the squared error loss function. The admissibility of estimators are discussed. Also, the problem of predicting the future record values, either point or interval, based on the past record values observed, is considered from Bayesian and non-Bayesian approaches.