Inference on the log-exponentiated Kumaraswamy distribution

Abstract

In this paper, the log-exponentiated Kumaraswamy (LEK) distribution is introduced and studied as a survival model of unemployment, its survived function has the interesting property that it can be decreasing depending on the shape parameters. The method of maximum likelihood is applied for estimating the model parameters, survival and hazard rate functions. Stratification is used to reduce heterogeneity in survival unemployment data. To achieve this aim, three models are considered. In the first model, unemployment experience for all working ages (15-60) is studied. In the second model, unemployment experience for ages (30-60) is studied and in the third model, unemployment experience for ages (45-60) is studied. Unemployment is modeled as a function of age. The distribution of unemployment with respect to age is represented by the LEK distribution or the three suggested models. For different values of samples sizes, Monte Carlo simulation is performed to investigate the precision of maximum likelihood estimates.

Comparisons of six different estimation methods for log-Kumaraswamy distribution

“…Some studies on LKw distribution can be listed as follows. Mohammed [3] studied inference on the log-exponentiated Kw distribution. Chacko and Mohan [4] investigated the problem of the estimation of parameters for Kw-exponential distribution under progressive type-II censoring.…”