A Proposed two-stage sequential point estimation of the hazard rate function

Abstract

We consider the problem of minimum risk point estimation of the exponential hazard rate. We assume that σ > σ_L, where σ_L > 0 is known to the experimenter from past experiences. A modified two-stage procedure is proposed which is shown to possess some first-order properties and is risk efficient. Simulation results are given and results indicate some small sample behaviour and provide support for the asymptotic behaviour of the sequential procedure as the cost c goes to zero. Hence, the proposed two-stage sequential procedure appears to be effective and useful.