Search

In this paper, genetic algorithm approach is used to estimate parameters of the Kumaraswamy distribution with maximum likelihood method. In addition ranked set sampling is used since it is expected to give better results in comparison to simple random sampling. Genetic algorithm approach is chosen because it is relatively more convenient in terms of satisfying positivity constraints for the parameters of the Kumaraswamy distribution. Also there is no need to use derivatives in the genetic algorithm approach. Bias, MSE and efficiency is calculated to compare performaces of maximum likelihood estimators for ranked set sampling and simple random sampling obtained by using genetic algorithms. The R software and related packages are preferred for calculations in the simulation study. [ABSTRACT FROM AUTHOR]