Open Access Open Access  Restricted Access Subscription Access

Less Restrictive Assumption for Distributional Characterization: A g-and-k Distribution Approach

Azmeri Khan, Sunzida Rahman, Syed Shahadat Hossain

Abstract



Whenever we choose to assume a specific distribution to be fitted to a given set of data, a wrong choice is always a possibility. The consequences of a incorrect assumption will lead to misleading inferences. Consideration of a less restrictive assumption by employing a more general and broader family of distribution may be thought to be one way for avoiding such consequences. In this paper, employment of a g-and-k distribution for the purpose is investigated. To check the adequacy of the use of g-and-k distribution as a underlying assumption on data, a simulation study has been conducted. Data are generated from six different theoretical continuous distributions: normal, logistic, laplace, half-normal, gamma and lognormal. For the generated datasets, both likelihood and Bayesian methods are used to characterize the distribution under both the assumptions that the data comes from the given parent distribution as well as from a g-and-k distribution. For comparison purpose the fitted quantiles of the data under the g-and-k distribution assumption against those under the specific parent distribution assumption are plotted. In the Bayesian set up, the twenty fifth, fiftieth and seventy fifth quantiles generated from the posterior distribution of the parameters using both approaches are compared. In all comparisons, the distribution
characterization under g-and-k distribution assumption performed almost similar to that under the true parent distribution assumption. Fitting real data employing the g-and-k distribution assumption is, that is why, deemed to be as good as fitting without having any rigid assumption, and as a result, is a robust approach in distribution characterization.

Keywords


g-and-k distribution, maximum likelihood method, posterior distribution, MCMC

Full Text:

PDF


Disclaimer/Regarding indexing issue:

We have provided the online access of all issues and papers to the indexing agencies (as given on journal web site). It’s depend on indexing agencies when, how and what manner they can index or not. Hence, we like to inform that on the basis of earlier indexing, we can’t predict the today or future indexing policy of third party (i.e. indexing agencies) as they have right to discontinue any journal at any time without prior information to the journal. So, please neither sends any question nor expects any answer from us on the behalf of third party i.e. indexing agencies.Hence, we will not issue any certificate or letter for indexing issue. Our role is just to provide the online access to them. So we do properly this and one can visit indexing agencies website to get the authentic information. Also: DOI is paid service which provided by a third party. We never mentioned that we go for this for our any journal. However, journal have no objection if author go directly for this paid DOI service.