Panuwat Pimsap. Alpha power exponentiated pareto family of distributions. Doctoral Degree(Statistics). Kasetsart University. Office of the University Library. : Kasetsart University, 2023.
Alpha power exponentiated pareto family of distributions
Abstract:
In this study, we propose distributions for the alpha power transformation family, encompassing the alpha power exponentiated Pareto distribution for lifetime data and the alpha power exponentiated generalized Pareto distribution for heavytailed data. We provide a comprehensive analysis of the mathematical properties of the proposed distribution, including the cumulative distribution function, probability density function, linear representation, moment generating function, moments, and order statistics. The simulation study evaluates the performance of parameter estimation using three methods: maximum likelihood, least squares, and Cramérvon Mises estimations. Furthermore, we present graphical representations of the probability density functions associated with the proposed distributions to illustrate their flexibility and suitability for modeling various data types. To assess the accuracy of parameter estimation, the root mean square error is selected as the criterion for comparing parameter estimation methods. In particular, the simulation results indicate that both the Cramérvon Mises and least squares estimations yield smaller errors compared to other estimation methods for both the alpha power exponentiated Pareto and the alpha power exponentiated generalized Pareto distributions. In addition, we apply our proposed distribution to three real data sets to demonstrate the effectiveness of fitting the proposed distribution to empirical data. Evaluation criteria, such as negative-log-likelihood, Akaike information criterion, Bayesian information criterion, and Kolmogorov-Smirnov statistics consistently indicate that the proposed distribution provides a better fit compared to existing distributions.
Kasetsart University. Office of the University Library