Estimation of the Exponential Distribution under Generalized Multiply Hybrid Censored Competing Risks Data

Abstract

The disadvantages of the multiply censoring scheme are that the experiment may be very long if the items are highly reliable. Because of that, multiply hybrid censoring scheme was introduced. But, one limitation of the multiply hybrid censoring scheme (MHCS) is that it cannot be applied when very few failures may occur before pre-fixed time. In this reason, we propose a generalized multiply hybrid censoring scheme, which allows us to observe a pre-specified number of failures. Also, we derive the maximum likelihood estimators and Bayes estimators of parameters based on exponential distribution with competing risks data under generalized multiply hybrid censoring scheme. Lindley’s approximation method (LAM) is used to compute these Bayes estimators. The Bayes estimators of parameters are better than the respective maximum likelihood estimators in terms of mean squared errors and biases. The choice of SEL seems to be a reasonable for Bayes estimation method of parameters. Also, a real data analysis has been provided.