Evolutionary estimation of parameters of Johnson distributions
Stefan Niermann · Journal of Statistical Computation and Simulation · 2006
The problem of fitting a Johnson distribution to data for situations in which the family membership is not known, a priori is considered in this article. Within each Johnson family member, the maximum likelihood estimator is determined with an evolutionary algorithm and the model with the overall highest likelihood is selected. A simulation experiment is performed to demonstrate the appropriateness of this new method.