|
Abstract
|
Continuous distributions have been recently much interest and developing dramatically. In this case, some of them has this ability to be applied instead of classical distribution to model real data. In this work, two distribution families, GOGa-G and EEG, are generated and introduced. To introduce them, the statistical properties of the distributions such as quantile function, asymptotics, moments, and entropy and some real dta set fiited is taken into account. Furthermore, the evaluation of some estimation methods such as Maximum Likelihood Estimation and Bayesian estimator is investigated by means of simulation. For this purpose, a specific package (in R software) and module (in JAGS software) dedicated just for this thesis will be applied. Most of the well-known distributions are the special type of the two new introduced family of distributions in this thesis, and due to the flexibility of the two families, they are able to be applied in variety of applications.
|