Affine-equivariant inference for multivariate location under Lp loss functions
针对Lp损失下的多元位置估计问题，提出一类基于数据单纯形的最小化估计量，实现仿射等变性，并证明其渐近正态性，在球对称下效率无损失，同时给出配套的假设检验方法。
We consider the fundamental problem of estimating the location of a d-variate probability measure under an Lp loss function. The naive estimator, that minimizes the usual empirical Lp risk, has a known asymptotic behavior but suffers from several deficiencies for p≠2, the most important one being the lack of equivariance under general affine transformations. In this work, we introduce a collection of Lp location estimators μˆnp,ℓ that minimize the size of suitable ℓ-dimensional data-based simplices. For ℓ=1, these estimators reduce to the naive ones, whereas, for ℓ=d, they are equivariant under affine transformations. Irrespective of ℓ, these estimators reduce to the sample mean for p=2, whereas for p=1, the estimators provide the well-known spatial median and Oja median for ℓ=1 and ℓ=d, respectively. Under very mild assumptions, we derive an explicit Bahadur representation result for μˆnp,ℓ and establish asymptotic normality. We prove that, quite remarkably, the asymptotic behavior of the estimators does not depend on ℓ under spherical symmetry, so that the affine equivariance for ℓ=d is achieved at no cost in terms of efficiency. To allow for large sample size n and/or large dimension d, we introduce a version of our estimators relying on incomplete U-statistics. Under a centro-symmetry assumption, we also define companion tests ϕnp,ℓ for the problem of testing the null hypothesis that the location μ of the underlying probability measure coincides with a given location μ0. For any p, affine invariance is achieved for ℓ=d. For any ℓ and p, we derive explicit expressions for the asymptotic power of these tests under contiguous local alternatives, which reveals that asymptotic relative efficiencies with respect to traditional parametric Gaussian procedures for hypothesis testing coincide with those obtained for point estimation. We illustrate finite-sample relevance of our asymptotic results through Monte Carlo exercises and also treat a real data example.