Comparisons of MVU and ML Distribution Function Estimators for Truncation Parameter Distributions
截断参数分布中 MVU 与 ML 分布函数估计量的比较
The ratio of integrated mean square errors for minimum variance unbiased and maximum likelihood estimating distribution functions is computed for truncation parameter distributions as a relative efficiency measure. A “closeness” relative efficiency is also computed for these estimators for the same class of distributions. The MVU estimator is found to be more efficient than the ML estimator for all sample sizes exceeding two in both cases.