Please use this identifier to cite or link to this item:
Title: Symmetric Generalized Two-Sided Power Distribution: Methods for Numerical Parameter Estimation
Authors: Raja’a Ali AL-Suraij 
Supervisor: Prof. Ahmad A. Soltani
Keywords: Two-Sided Power Distribution;Numerical Parameter Estimation
Issue Date: 2015
Publisher:  Kuwait university - college of graduate studies
Abstract: Symmetric generalized two-sided power distribution are introduced by Soltani and Homei (2009), where the density function formulation is presented in terms of Euler’s integral representation of the Gauss hypergeometric function. Indeed, this distribution belongs to a very special and interesting class of distributions. A symmetric generalized two-sided power distribution assumes a normal shaped density function on finite interval, and takes a value of zero on the interval end points. In many applications, variables under investigation assume values on a finite interval, and exhibit a more or less symmetric, normal shaped histogram. In this thesis, we focus on making statistical inferences and simulations for the symmetric generalized two-sided power distribution. Not only are these topics important in practice, but they are also of theoretical interest. Due to the complications in the density formulation, it is not possible to apply the classical estimation methods. For the inference, we examine three numerical estimation procedures: (i) the classical maximum likelihood estimation (MLE), (ii) the method of moments (MM) and (iii) the adjusted method of moments (AMM). In this thesis we put light on the effectiveness of the adjusted method of moments estimation procedure. We support this investigation with numerous simulated data sets.
Appears in Programs:0480 Statistics & Operations Research

Files in This Item:
File Description SizeFormat 
Rajaa_Ali_Thesis_June_2015.pdf398,42 kBAdobe PDFView/Open    Request a copy
Show full item record

Page view(s) 20

Last Week
Last month
checked on Feb 27, 2020


checked on Feb 27, 2020

Google ScholarTM


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.