位置刻度参数最优同变区间估计的minimax性质

Let X be a p-dimensional random vector having a density function σ^-p((x-θ1)/σ),where p≥2, x∈R^p, θ∈R¹, σ>0, 1= (1, …, 1)_p×1¹. The minimax property of best equivariant confidence intervals for the location parameter θ and scale parameter σ are proved under followiny loss function

On the Minimax Property of Best Equivariant Confidence Intervals for the Location and Scale Parameters