1. Department of Mathematics, Indian Institute of Technology Patna, Bihar, India;2. National Institute of Pharmaceutical Education and Research, Hajipur, Bihar, India;3. Department of Statistics, University of Mazandaran, Babolsar, Iran
Abstract:
We consider estimation of unknown parameters of a Burr XII distribution based on progressively Type I hybrid censored data. The maximum likelihood estimates are obtained using an expectation maximization algorithm. Asymptotic interval estimates are constructed from the Fisher information matrix. We obtain Bayes estimates under the squared error loss function using the Lindley method and Metropolis–Hastings algorithm. The predictive estimates of censored observations are obtained and the corresponding prediction intervals are also constructed. We compare the performance of the different methods using simulations. Two real datasets have been analyzed for illustrative purposes.