A0823
Title: Operationally convenient and risk efficient estimation of the mean exponential lifetime under interval censoring
Authors: Jun Hu - Oakland University (United States) [presenting]
Abstract: A novel two-stage sampling scheme is introduced for the minimum risk point estimation (MRPE) of the mean exponential lifetime under interval censoring, addressing the limitations of existing risk-efficient sequential procedures that require continuous monitoring and real-time detection of failures and are often operationally inconvenient. The proposed procedure involves a first-stage sample size determined by a lower bound of the optimal sample size, followed by a second-stage sample size based on data-driven adjustment. The asymptotic efficiency and risk efficiency properties are established through theoretical proofs and extensive Monte Carlo simulations. To illustrate the practical applicability of the procedure, a real-world dataset is analyzed on remission times for bladder cancer patients, demonstrating its ability to efficiently estimate the exponential mean with a reduced sample size.
