Published May 3, 2022 | Version v21
Working paper Open

Weibull, κ-Weibull and other probability distributions

  • 1. Politecnico di Torino

Description

Here we will consider a function of κ-statistics, the κ-Weibull distribution, and  compare it to the well-known Weibull distribution. The κ-Weibull will be also compared to  the 3-parameter extended Weibull function, obtained according to the Marshall–Olkin extended distributions. The log-logistic distribution will be considered for comparison too, such as the exponentiated Weibull, the Burr and the q-Weibull distributions. The most important observation, coming from the proposed calculations, is that the κ-Weibull hazard function is strongly depending on the values of parameter  κ, a parameter which is deeply influencing the behaviour of the tail of the probability distribution. As a consequence, the  κ-Weibull function turns out to be quite relevant for generalizations of the Weibull approach to modeling failure times. Discussions about the Maximum Likelihood approach for Weibull, κ-Weibull and Burr distributions will be also given.

Files

weibull20.pdf

Files (1.5 MB)

Name Size Download all
md5:d61a8f46aa79f61c893ca276bfbc2cd1
1.5 MB Preview Download