Bayesian Estimation for a Finite Mixed Exponential Distribution
Ping Chen · Journal of Gansu Sciences · 2008
The mixed exponential distribution plays an important part in lifetime data analysis but we may suffer from extreme difficulties to get the maximal likelihood estimate due to the complexity of likelihood function,ZHU Liping etc.(2006) employed the EM algorithm to study parameters estimation of a mixture model of exponential distributions and obtained good results.We study parameters estimation of a finite mixture model of exponential distributions with two parameters under full data and censored data by means of Gibbs sampler algorithm in Bayesian framework.The simulations suggest that the Baysian algorithm is very effective to our model.