The effect of polluted samples on Bayesian Estimators of Burr type –XII distribution

Abstract

Bayesian estimators may be affected by the polluted samples, because these samples can lead to the influence of the estimation methods in general and the Bayesian methods in particular, and thus the deviation of the values of the distribution parameter from their real values, and this leads to the divergence of the capabilities of the Bayes survival estimators from the real values.
The results showed that the estimators of the parameters were affected by many factors (sample size, distribution parameter, number of outliers and the estimation method). Simulation experiment results also showed a difference in Mean Square Error (MSE) of the Bayes survival estimators for each different experiment. Bayesian methods can be compared with other estimation methods (Maximum likelihood Estimation (MLE), Moment estimation (MOM) and shrinkage method (SH)). Also, Bayesian methods can be used to estimate the survival function of other distributions (exponential, Gamma and mixed) to observe the estimation results with the presence of extreme values.