鞅型序列的变换及其收敛性

本文证明了(1)设 Banach 空间 B 为 P 阶光滑的(1≤P≤2),X=(X_n,(?)_n,n≥1)为B 值鞅,v=(v_n,(?)_n,n≥1)为实值可予报序列,鞅变换 Y=(sum from i=1 to n V_i(X_i-X_(i-1)),(?)_n,n≥1)在一定的条件下具有 a.e.收敛性,L~p 收敛性及强(弱)大多数定律成立。(2)Banach空间 B 具有 Radon-Nikodym 性质,X=(X_n,(?)_n,n≥1)为 B 值依概极限鞅,实值可予报序列 V=(V_n,(?)_n,n≥1)满足 sum from i=1 to ∞ E(|V_i|~p)~(1/p)<∝,1相似文献   

线性回归系数两种可估性的关系

<正> 设有满足Gauss-Markov(GM)条件的线性模型Y=Xβ+e,E_e=0,COV(e)=б~2l,此处X为已知的n×p矩阵,β=(β_1,…,β_p)′为p维未知向量,I为n阶单位阵,0<σ~2<∞,σ~2也未知.设c为一已知的p维向量,则当x的秩为P时,c′β必为线性可估.反过来,若X的秩小于p,则对某些c,c′β不是线性可估,甚至也可以不是可估的.     

相依误差下回归系数 LS 估计的收敛速度

其中 x(ij)(j=1,…,p,i=1,2,…)是已知常数,常称之为模型(1)的设计常数或设计点列,β_1,…,β_p,为未知的回归系数,y_i,e_i 分别为第 i 次量测时的量测值和量测随机误差。以下,我们记设计矩阵(x_(ij))(?)≤(?)≤n,(?)≤j p 为 X_n,Y_n=(y_1,…,y_n)′,β=(β_1,…,β_p)′.并假定对某N,X′_N X_N 非退化,那么当 n≥N 时,X′_n X_n 亦非退化,且回归系数β的基于前 n 次量测值 Y_(?)及设计矩阵 X_(?)的最小二乘估计(通常简记为 LS 估计) b_(?)=(b_(?)1, …,b_((?)p)'为     

线性模型中误差方差估计的非一致性估计

<正> 考虑线性模型 Y_((n))=X_nβ+e_((n)),n=1,2,…其中X_n=(xij)为n×p阶已知矩阵,rank X_n=r_n,为未知的待估参数向量,e_((n))=为独立的试验误差,满足条件     

矩阵损失下回归系数和误差方差的联合估计的可容许性的几个结果

设有线性模型:Y=(Y_1,…,Y_n)＇=Xβ+ε=Xβ+(ε_1,…,ε_n)＇,其中X:n×p已知,β=(β_1,…,β_p)＇未知,ε_1,…,ε_n独立,E_(ε_i)=E_(ε_i~3)=0,E_(ε_4~2)=σ~2,F_(ε_i~4)=3σ~4,i=1,2,…,n,0<σ~2<∞,σ~2未知。在矩阵损失下,我们考虑(Sβ,σ~2)的联合估计(AY,Y＇BY)在估计类×={(CY,Y＇DY):C为m×n的常数阵,D≥0为n×n的常数阵中的可容许性,得到了(AY,Y＇BY)为(Sβ,σ~2)的可容许估计的一些充分条件和必要条件。     

强收敛意义下岭回归的 C_L 和 GCV 准则的渐近最优性

设有回归模型Y_i=μ_i+e_i,i=1,2,…,n (1)假定 e_1,…,e_n 为 iid.的正态随机变量序列,具有共同的均值0和方差σ~2.每个 Y_i 可通过设计点列 x_(i1),x_(i2),…,x_i_p_n 观察到.为估计 Y=(Y_1,…,Y_n)′的未知均值 μ=(μ_1,…,μ_n)′,可构造一族岭估计(?)(h)=X(X′X+hI)~-1X′Y,h≥0,(2)其中 X=(x_ij)_(n×ρn) 为设计阵,I 为 p_n 阶单位阵.在这里,岭参数 h 的选择是一个十分     

线性模型中误差方差估计相合性的收敛速度

考虑线性模型y_j=x′β+e_j,j=1,2,…,n,假定误差序列{e_j}为ⅱd,随机变量序列,满足Ee_1=0,00,n→∞; (ⅱ)对任何ε>0,n→∞; (ⅲ) 这个结果与{Y_j}为独立同分布场合完全一致。     

非参数回归函数最近邻估计的强收敛速度

<正> §1.引言 设(X,Y),(X_1,Y_1),…,(X_n,Y_n)为iid d×1维随机向量,E|Y|<∞.对x=(x~(1)),…,x~(d))∈R~d,取‖x‖为欧氏模或对固定的x∈R~d,将(X_1,Y_1),…,(X_n,Y_n)按照     

重特征值敏度的数值计算

一个结构系统的设计,往往归结为下述代数特征值问题:其中A(p)与B(p)为n×n实解析的对称矩阵,B(p)正定,λ(p)是特征值,x(p)是相应的特征向量. 设λ_1是问题(1.1)在点p=p~*的r重特征值,即存在矩阵X=(X_1,X_2)∈R~(n×n),     

代数特征值反问题之解的稳定性分析

考虑如下代数特征值反问题: 问题 G(A;{A_k}_1~n;λ).设 A=(a_(ij)),A_k=(a_(ij)~((k))),k=1,…,n是n+1个n×n的实对称矩阵,λ=(λ_1,…,λ_n)是n维实向量且λ_i≠λ_j,i≠j.求n维实向量c=(c_1,…,c_n)~T,使矩阵A(c)=A+sum from k=1 to n (c_kA_k)的特征值是λ_1,…,λ_n. 这一问题是经典加法问题的推广.当A_k-e_ke_k~~T(e_k是n阶单位阵的第k列)时,     

Symmetrized approximate score rank tests for the two-sample case

Rank test statistics for the two-sample problem are based on the sum of the rank scores from either sample. However, a critical difference can occur when approximate scores are used since the sum of the rank scores from sample 1 is not equal to minus the sum of the rank scores from sample 2. By centering and scaling as described in Hajek and Sidak (1967, Theory of Rank Tests, Academic Press, New York) for the uncensored data case the statistics computed from each sample become identical. However such symmetrized approximate scores rank statistics have not been proposed in the censored data case. We propose a statistic that treats the two approximate scores rank statistics in a symmetric manner. Under equal censoring distributions the symmetric rank tests are efficient when the score function corresponds to the underlying model distribution. For unequal censoring distributions we derive a useable expression for the asymptotic variance of our symmetric rank statistics.     

椭球检验与统计量的渐近尾部性

In this paper,some test statistics of Kolmogorov type and Cramervon Mises type based on projection pursuit technique are proposed for testing the sphericity problem of a high-dimensional distribution. The limiting distributions of the test statistics are derived under the null hypothesis. The asymptotic properties of Bootstrap approximation are investigated and the tail behaviors of the statistics are studied.     

缺失数据下非线性EV模型参数的经验似然置信域

考虑解释变量带有测量误差且响应变量随机缺失情形下的非线性EV模型.通过利用核实数据, 构造了未知参数的两种经验对数似然比统计量. 证明了所构造统计量的分布渐近于χ²分布, 所得结果可以用来构造未知参数的渐近置信域.     

Tests of elliptical symmetry and the asymptotic tail behavior of the statistics

In this paper, some test statistics Of Kolmogorov type and Cramervon Mises type based on projection pursuit technique are proposed for testing the sphericity problem of a high-dimensional distribution. The limiting distributions of the test statistics are derived under the null hypothesis. The asymptotic properties of Bootstrap approximation are investigated and the tail behaviors of the statistics are studied.     

On the asymptotic simultaneous distribution of randomly indexed order statistics

The asymptotic simultaneous distribution of normalized central order statistics for the random size of sample is studied. This study develops works [1, 2], in which Student’s distribution acts as the limit distribution for a class of statistics. Results from the study are applied to construct statistical insights on the shift/scale parameter ratio for two-parametric families of distributions.     

Average projection type weighted Cramér-von Mises statistics for testing some distributions

This paper addresses the problem of testing goodness-of-fit for several important multivariate distributions: (Ⅰ) Uniform distribution on p-dimensional unit sphere; (Ⅱ) multivariate standard normal distribution; and (Ⅲ) multivariate normal distribution with unknown mean vector and covariance matrix. The average projection type weighted Cramér-yon Mises test statistic as well as estimated and weighted Cramér-von Mises statistics for testing distributions (Ⅰ), (Ⅱ) and (Ⅲ) are constructed via integrating projection direction on the unit sphere, and the asymptotic distributions and the expansions of those test statistics under the null hypothesis are also obtained. Furthermore, the approach of this paper can be applied to testing goodness-of-fit for elliptically contoured distributions.     

ON SOME TESTS FOR ELLIPTICAL SYMMETRY

In this paper, some test statistics of Kolmogorov type and Cramer-von Mises type based on projection pursuit technique are proposed for testing the sphericity problem of a high-dimensional distribution. The limiting distributions of the test statistics are derived under the null hypothesis and any fixed alternative. The asymptotic properties of Bootstrap approximation are investigated. Furthermore, for computational reasons, an approximation for the statistics, based on number theoretic method, is suggested.     

Quantitative Approximation to the Ordered Dirichlet Distribution under Varying Basic Probability Spaces

An approximate expansion of a sequence of ordered Dirichlet densities is given under the set-up with varying dimensions of the relating basic probability spaces. The problem is handled as the approximation to the joint distribution of an increasing number of selected order statistics based on the random sample drawn from the uniform distribution U(0, 1). Some inverse factorial series to the expansion of logarithmic function enable us to give quantitative error evaluations to our problem. With the help of them the relating modified K-L information number, which is defined on an approximate main domain and different from the usual ones, is accurately evaluated. Further, the proof of the approximate joint normality of the selected order statistics is more systematically presented than those given in existing works. Concerning the approximate normality the modified affinity and the half variation distance are also evaluated.     

On goodness-of-fit tests for multiple recursive random number generators

This paper considers the problem of reducing the computational time in testing uniformity for a full period multiple recursive generator (MRG). If a sequence of random numbers generated by a MRG is divided into even number of segments, say 2s, then the multinomial goodness-of-fit tests and the empirical distribution function goodness-of-fit tests calculated from the ith segment are the same as those of the (s + i)th segment. The equivalence properties of the goodness-of-fit test statistics for a MRG and its associated reverse and additive inverse MRGs are also discussed.     

Average projection type weighted Cramér- von Mises statistics for testing some distributions

This paper addresses the problem of testing goodness-of-fit for several important multivariate distributions: (I) Uniform distribution on p-dimensional unit sphere; (II) multivariate standard normal distribution; and (III) multivariate normal distribution with unknown mean vector and covariance matrix. The average projection type weighted Cramér-von Mises test statistic as well as estimated and weighted Cramér-von Mises statistics for testing distributions (I), (II) and (III) are constructed via integrating projection direction on the unit sphere, and the asymptotic distributions and the expansions of those test statistics under the null hypothesis are also obtained. Furthermore, the approach of this paper can be applied to testing goodness-of-fit for elliptically contoured distributions.