Advances and Applications in Statistics
Volume 58, Issue 1, Pages 57 - 75
(September 2019) http://dx.doi.org/10.17654/AS058010057 |
|
USE OF EXPONENTIAL DISTRIBUTION FOR HYBRIDIZATION OF DISTRIBUTIONS
Dina H. Abdel Hady
|
Abstract: Statistical distributions are beneficial in predicting and depicting the phenomena of the real world. Though many distributions were developed, there are always scopes for improving distributions that are more flexible. Afify [2] introduced a new technique called hybridization distributions based on proportional hazard model (PHM) and rank transmutation map (RTM). In this paper, we introduce methods for developing distributions which are new and more flexible. These methods come from hybridization proportional hazard model (PHM) and proportional reversed hazard model (PRHM). In addition, we study the properties of these methods and compare it with the old models in case of complete sample, Type I and Type II censored sample using Kolmogorov-Smirnov (K-S) statistic for goodness of fit. Finally, comparisons between the maximum likelihood using Monte Carlo simulation and Bayes estimators are conducted using MCMC study. |
Keywords and phrases: proportional hazard method, proportional reversed hazard method, exponential distribution, likelihood function, Bayes estimators, the quartiles, order statistics, censored data, Monte Carlo and MCMC simulations.
|
|
Number of Downloads: 386 | Number of Views: 948 |
|