Bayesian Analysis of Zero-Inflated Generalized Power Series Distributions Under Different Loss Functions
The values of the expected frequencies clearly shows that the zero-inflated Poisson distribution provided a closer fit than that provided by the classical Poisson distribution. It is also clear from the table that the Bayes estimators obtained under weighted squared error loss functions (WSELF) gives closer fits than the Bayes estimator obtained under squared error loss function (SELF). Also, the exponentially minimum expected loss (EWMEL) estimates gives better fits than the minimum expected loss (MEL) estimates. Keeping in view the importance of count data modeling it is recommended that whenever the experimental number of zeros are more than that given by the model, the model should be adjusted accordingly to account for the extra zeros.
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.