On an Extension of the Exponentiated Weibull Distribution |
| |
Authors: | H Bidram M H Alamatsaz V Nekoukhou |
| |
Institution: | 1. Department of Statistics, University of Isfahan, Isfahan, Iran;2. Department of Statistics, University of Isfahan, Khansar Unit, Isfahan, Iran |
| |
Abstract: | In this article, the exponentiated Weibull distribution is extended by the Marshall-Olkin family. Our new four-parameter family has a hazard rate function with various desired shapes depending on the choice of its parameters and, thus, it is very flexible in data modeling. It also contains two mixed distributions with applications to series and parallel systems in reliability and also contains several previously known lifetime distributions. We shall study some basic distributional properties of the new distribution. Some closed forms are derived for its moment generating function and moments as well as moments of its order statistics. The model parameters are estimated by the maximum likelihood method. The stress–strength parameter and its estimation are also investigated. Finally, an application of the new model is illustrated using two real datasets. |
| |
Keywords: | Beta-generalized (exponentiated) Weibull distribution Compounding Marshall–Olkin family Maximum likelihood estimation Mixing Stress–strength parameter |
|
|