
Bayes Estimators for the Parameter of the Inverted Exponential Distribution Under different Double informative priors
Author(s) -
جنان عباس ناصر
Publication year - 2018
Publication title -
mağallaẗ al-ʿulūm al-iqtiṣādiyyaẗ wa-al-idāriyyaẗ
Language(s) - English
Resource type - Journals
eISSN - 2518-5764
pISSN - 2227-703X
DOI - 10.33095/jeas.v24i103.134
Subject(s) - prior probability , estimator , bayes' theorem , mathematics , exponential distribution , statistics , distribution (mathematics) , bayesian probability , gamma distribution , bayes' rule , exponential function , econometrics , computer science , artificial intelligence , bayes factor , mathematical analysis
In this paper, we present a comparison of double informative priors which are assumed for the parameter of inverted exponential distribution.To estimate the parameter of inverted exponential distribution by using Bayes estimation ,will be used two different kind of information in the Bayes estimation; two different priors have been selected for the parameter of inverted exponential distribution. Also assumed Chi-squared - Gamma distribution, Chi-squared - Erlang distribution, and- Gamma- Erlang distribution as double priors. The results are the derivations of these estimators under the squared error loss function with three different double priors.
Additionally Maximum likelihood estimation method (MLE) was used to estimate the parameter of inverted exponential distribution .We used simulation technique, to compare the performance for each estimator, several cases from inverted exponential distribution for data generating, for different samples sizes (small, medium, and large).Simulation results shown that the best method is the bayes estimation according to the smallest values of mean square errors( MSE) for all samples sizes (n) comparative to the estimated values by using MLE . According to obtained results, we see that when the double prior distribution for is Gamma- Erlang distribution for some values for the parameters a, b & given the best results according to the smallest values of mean square errors (MSE) comparative to the same values which obtained by using Maximum likelihood estimation (MLE) for the assuming true values for and for all samples sizes.