Risk-Efficient Estimation of the Mean Exponential Survival Time under Random Censoring

摘要

The paper proposes a sequential estimator theta of the parameter theta of an exponential distribution when the data is censored. Without any further conditions, it is shown that theta is asymptotically risk efficient when the loss is measured by the squared error loss of estimation of theta plus a linear cost function of the number of observations. In addition, it is shown that theta is asymptotically normal as the cost per observation to zero. (Author)