ON FINE-TUNED BOUNDED RISK SEQUENTIAL POINT ESTIMATION OF THE MEAN OF AN EXPONENTIAL-DISTRIBUTION

The bounded risk point estimation problems for the mean of a one-parameter exponential population are addressed under the squared error loss function. Both fully sequential as well as accelerated sequential methodologies are utilized and the second-order approximations to the associated risk functions are obtained. The idea of fine-tuning these sampling strategies is pursued so that the associated risk functions can be expanded as the sum of the nominal preassigned bound and a remainder term which is shown to have the same o(.) order as that obtained by the original fully sequential process. The explicit expressions of these fine-tuning factors have been provided through our analysis.