分块奇异线性模型及其导出的奇异线性模型间的最小范数二次无偏估计等价性研究

本文考虑一般线性模型A=(y,X1β1 X2β2,σ^2V)及其导出线性模型，其中V是已知的非负定矩阵，X=(X1：X2)是已知的设计矩阵，给出了线性模型A及其导出线性模型间最小范数二次无偏估计间差的表达式，更进一步，建立了线性模型A及其导出线性模型间最小范数二次无偏估计相等的充分必要条件。     

一般线性模型下方差最小范数二次无偏估计的稳健性

本文给出了一般线性模型下,由最小二乘得到的方差估计与最小范数二次无偏估计相等的充分必要条件,并且当Gauss-Markov估计与最小二乘估计相等时,可以得到一个简单的等价条件。     

随机效应模型中方差分量的非负二次型估计

设Y_1,Y_2是相互独立的随机变量,Y_1/(ασ+τ)～X~2(n_1),Y_2/τ～x~2(n_2),其中σ>0,τ>0是未知方差分量,α>0,正整数n_1,n_2是已知常数。本文从风险函数及偏差角度研究了σ的无偏估计Y_1/(αn_1)—Y_2/(αn_2)的改进,并指出用非负二次估计PY_1代替σ的无偏估计较合适,其中最后把上述结果用于一种方式分组及二级套分类随机效应模型。     

正态—平稳二项过程模型结构可靠度的最小方差无偏估计

本文讨论强度随机变量Ｘ服从正态分布,应力｛Ｙ（ｔ）,ｔ∈［０,Ｔ］｝为平稳二项随机过程模型结构可靠度的最小方差无偏估计．     

二次损失下回归系数的线性Minimax估计

设有线性模型 EY=Xβ CovY=σ~2V, 这里X: _(nxp),和V: _(nxn)>0已知矩阵,β∈R~P和σ~2>0都是参数。本文估计Sβ,选取损失函数 L(d,Sβ)=((d-Sβ)′(d-Sβ))/(σ~2+β′X′V~(-1)Xβ), 其中Sβ是可估的,并给出了在线性估计类中唯一的一个线性minimax估计。     

方差分量的二次不变估计的非负改进

研究一类方差分量模型中的方差分量的估计改进问题,首先在含两个方差分量模型中给出σ21二次型估计类,并且此估计类还具有无偏性和不变性.考虑二次损失(δ-θ)2,在此估计类基础上放弃无偏性进行非负改进,不仅得到优于二次不变无偏估计类的σ21的非负二次不变估计类,而且还说明了它优于方差分析估计和最小均方误差估计,文献[5]中给出s>2时的非负改进,但是非负改进存在是有条件的,本文克服了这个缺陷.最后给出了非负改进存在的充分必要条件.     

线性模型中均值向量的最小二乘估计和最佳线性无偏估计之差的范数界的注记

我们讨论一般线性模型:Y=Xβ e,E(e)=0,Cov(e)=σ~2V,V为非负定协方差矩阵。我们知道μ=Xβ的最小二乘估计和最佳线性无偏估计分别为μ~*=X(X′X)~-X′Y和■=X(X′T~-X)~-X′X~-Y,这里T=V XUX′,U是一个对称阵使得R(T)=R(V■X)以及T≥0。本文讨论V≥0时,■与μ~*之差的范数界,把V>0时■和μ~*之差在Haberman条件下的范数界推广到V≥0,且在取常用的欧氏范数时,得到使Haberman条件成立的便于应用的充要条件。本文还证明了[2]界的推广形式,并把[3]界推广到V≥0的情况。     

平衡线性混合模型方差分量几种估计的优良性

对于平衡线性混合模型,本文提出了一组易验证的条件,在此条件下,方差分量的谱分解估计、方差分析估计和最小范数二次无偏估计都相等且为一致最小方差无偏估计.同时证明了在此条件下,似然方程和限制似然方程都有显式解,还给出了许多满足这组条件的平衡线性混合模型的例子.     

关于无偏估计的若干问题

通过理论分析和实例讨论有关无偏估计的若干问题：无偏估计不一定存在；无偏估计一般不唯一；在均方误差意义下，无偏估计不一定优于有偏估计；均方误差最小的估计是最小方差无偏估计，但反之不然。     

平衡线性混合模型方差分量几种估计的优良性

对于平衡线性混合模型,本文提出了一组易验证的条件,在此条件下,方差分量的谱分解估计、方 差分析估计和最小范数二次无偏估计都相等且为一致最小方差无偏估计．同时证明了在此条件下,似然 方程和限制似然方程都有显式解,还给出了许多满足这组条件的平衡线性混合模型的例子．     

All Admissible Linear Estimators under Quadratic Loss in Multivariate Model

§ 1.Introduction and Notations　In this paper,for any given matrices A and B,A B denotes the Kronecker productof A and B,A is a vector formed by stacking the columns of A under each other,μ(A)is a space generated by the columns of A,and PA=A(A′A) - A′. Fourthmore,if A andB are square matrices,then A>B and A≥ B mean that A-B is a symmetrical positiveand nonnegative matrix,respectively,andλi(A) is the i-th largest eigenvalue of A.　Consider general multivariate linear modelY …     

Estimators of covariances in time series models

The theory of Minimum Norm Quadratic Estimators for estimating variances and covariances is applied to show that some commonly used estimators of covariances in time series models are easily derived using the above principle. In applying the theory MINQE, it is observed that no unbiased estimator exists in the class of invariant quadratics.     

任意秩有限总体中无偏预测和估计的构造关系

研究了有限总体均值向量的无偏估计和线性可预测变量的无偏预测之间的关系,利用分块矩阵广义逆直接对加权风险函数进行分解,提出了一种由均值向量的无偏估计来构造无偏预测的新方法,并找到了它们之间的构造关系.特别地,线性可预测变量的最优线性无偏预测(BLUP)可由均值向量的最佳线性无偏估计(BLUE)惟一地表示(有关惟一性在几乎处处意义下理解).     

均匀分布参数的无偏估计及其分布

讨论了均匀分布未知参数无偏估计量的分布密度,利用无偏估计量构造出一些新的样本函数,并且利用给出的样本函数推导出了未知参数的置信区间.所得到结果改善了现有的估计,易于计算.     

A Comparison of Restricted and Unrestricted Estimators in Estimating Linear Functions of Ordered Scale Parameters of Two Gamma Distributions

The problem of estimating linear functions of ordered scale parameters of two Gamma distributions is considered. A necessary and sufficient condition on the ratio of two coefficients is given for the maximum likelihood estimator (MLE) to dominate the crude unbiased estimator (UE) in terms of mean square error. A modified MLE which satisfies the restriction is also suggested, and a necessary and sufficient condition is also given for it to dominate the admissible estimator based solely on one sample. The estimation of linear functions of variances in two sample problem and also of variance components in a one-way random effect model is mentioned.     

Minimum Divergence Estimators Based on Grouped Data

The paper considers statistical models with real-valued observations i.i.d. by F(x, ₀) from a family of distribution functions (F(x, ); ), R ^s , s 1. For random quantizations defined by sample quantiles (F _n ^–1 (₁),, F _n ^–1 ( _m–1)) of arbitrary fixed orders 0 < ₁ < _m-1 < 1, there are studied estimators _,n of ₀ which minimize -divergences of the theoretical and empirical probabilities. Under an appropriate regularity, all these estimators are shown to be as efficient (first order, in the sense of Rao) as the MLE in the model quantified nonrandomly by (F ^–1 (₁,₀),, F ^–1 ( _m–1, ₀)). Moreover, the Fisher information matrix I _m (₀, ) of the latter model with the equidistant orders = ( _j = j/m : 1 j m – 1) arbitrarily closely approximates the Fisher information J(₀) of the original model when m is appropriately large. Thus the random binning by a large number of quantiles of equidistant orders leads to appropriate estimates of the above considered type.     

A Subsampling Method for the Computation of Multivariate Estimators with High Breakdown Point

Abstract

All known robust location and scale estimators with high breakdown point for multivariate samples are very expensive to compute. In practice, this computation has to be carried out using an approximate subsampling procedure. In this article we describe an alternative subsampling scheme, applicable to both the Stahel-Donoho estimator and the minimum volume ellipsoid estimator, with the property that the number of subsamples required can be substantially reduced with respect to the standard subsampling procedures used in both cases. We also discuss some bias and variability properties of the estimator obtained from the proposed subsampling process.