a classical approach for parameters estimation of the exponentiated pareto distribution with type-ii censoring

in this paper we have taken into consideration the estimation of the exponentiated pareto distribution parameters when data are observed under type-ii censoring. maximum likelihood estimates are obtained using the newton-raphson and em algorithms. we also construct shrinkage preliminary test estimates to improve the estimation accuracy. it is observed that the exact biases and mean risks of the proposed estimates cannot be obtained in closed form; because of this, we use a mont carlo simulation study to assess performance of the proposed estimates.