Abstract:
This research proposed a Bayesian parameters estimation of mixture
generalized gamma distribution based on the loss functions. In this research, we
proposed Bayesian parameters estimation by simulation study for right skewed
mixture generalized gamma distribution with sample sizes of 20, 50, 100, 300.
Parameters estimation using Bayesian estimation under the loss function, which
consisted of squared error loss function, absolute error loss function, weighted
squared error loss function, precautionary loss function and k-loss function. The
performance of the Bayesian estimator was compared with posterior risk criteria, the
performance of Bayesian parameter estimation was compared with Maximum
likelihood estimation by mean squared error. The results show that the Bayesian
method gave less mean squared error, so it was concluded that the Bayesian
method performed better than Maximum likelihood method. The result will be
shown in application data the active repair times for an airborne communication
transceiver. The Bayesian method under the precautionary loss function is the most
appropriate estimate distribution model as it has the goodness of fit with data. This
will be useful in forecasting or policy maintenance of airborne communication
transceiver