Selection effect of learning rate parameter on estimators of k exponential populations under the joint hybrid censoring

Abstract

A Bayesian method based on the learning rate parameter η is called a generalized Bayesian method. In this study, joint hybrid censored type I and type II samples from k exponential populations were examined to determine the influence of the parameter η on the estimation results. To investigate the selection effects of the learning rate and the loss parameters on the estimation results, we considered two additional loss functions in the Bayesian approach: the linear and the generalized entropy loss functions. We then compared the generalized Bayesian algorithm with the traditional Bayesian algorithm. We performed Monte Carlo simulations to compare the performance of the estimation results with the losses and different values of η. The effects of different losses with different values and learning rate parameters are examined using an example.

Keywords

• Generalized Bayes
• Learning rate parameter
• Exponential distribution
• Joint hybrid censoring
• Linex loss
• General entropy loss