A Choice Between Poisson and Geometric Distributions

Article Type

Research Article

Publication Title

Journal of the Indian Society for Probability and Statistics

Abstract

Both Poisson and geometric distributions can be used quite effectively in modeling count data. In this paper we compare Bayesian and frequentists criteria to choose between Poisson and geometric distributions. In the frequentist choice, we use the ratio of maximized likelihood in discriminating between the two distributions. Asymptotic distributions of the ratio of maximized likelihoods are also obtained, and they can be used to compute the minimum sample size needed to discriminate between the two distributions. We further investigate the Bayesian model selection criterion in choosing the correct model, under a fairly general set of priors. We perform some simulation experiments to see the effectiveness of the proposed methods and the analysis of one data set is performed for illustrative purposes.

First Page

111

Last Page

123

DOI

10.1007/s41096-016-0008-2

Publication Date

12-1-2016

Link to Full Text

Share

COinS