Inadmissibility of Linearly Invariant Estimators in Truncated Parameter Spaces
研究了实际参数空间为全参数空间严格子集时的估计问题，在二次损失函数下证明靠近参数空间边界的估计量不可容许，并给出更优估计量，应用于不等式约束回归和随机化回答。
Abstract Estimation problems are considered where the actual parameter space is a strict subset of the full parameter space. A linear invariance structure is assumed and a quadratic loss function is used. Then it is shown that estimators taking values near the boundary of the parameter space are inadmissible and better estimators are presented. Applications are given in the fields of inequality constraint regression and randomized response.