共查询到20条相似文献,搜索用时 328 毫秒
1.
LU KUNLIANG 《系统科学与数学》1983,3(2):095-111
1. INTRODUCTION Consider the multiple regression model xi=ciβ+ei,i=1,2,…,(1) where c_i=(c_(il),…,c_(ip)), i=1, 2,…are given row vectors; β∈R~p is an unknown parameter vector; e_1, e_2,…are i.i.d. random variables with a common probability density function f(x) with respect to the Lebesgue measure. Let x_1, x_2,…be a sequence of observa- 相似文献
2.
For complete observation and p-dimensional parameterθdefined by an estimation equation,empirical likelihood method of construction of confidence region is based on the asymptoticχ2pdistribution of-2 log(EL ratio).For right censored lifetime data with covariables,however,it is shown in literature that-2 log(EL ratio)converges weakly to a scaledχ2pdistribution,where the scale parameter is a function of unknown asymptotic covariance matrix.The construction of confidence region requires estimation of this scale parameter.In this paper,by using influence functions in the estimating equation,we show that-2 log(EL ratio)converges weakly to a standardχ2pdistribution and hence eliminates the procedure of estimating the scale parameter. 相似文献
3.
In the receiver operating characteristic (ROC) analysis,the area under the ROC curve (AUC) is a popular summary index of discriminatory accuracy of a diagnostic test.Incorporating covariates into ROC analysis can improve the diagnostic accuracy of the test.Regression model for the AUC is a tool to evaluate the effects of the covariates on the diagnostic accuracy.In this paper,empirical likelihood (EL) method is proposed for the AUC regression model.For the regression parameter vector,it can be shown that the asymptotic distribution of its EL ratio statistic is a weighted sum of independent chi-square distributions.Confidence regions are constructed for the parameter vector based on the newly developed empirical likelihood theorem,as well as for the covariate-specific AUC.Simulation studies were conducted to compare the relative performance of the proposed EL-based methods with the existing method in AUC regression.Finally,the proposed methods are illustrated with a real data set. 相似文献
4.
ChaiGenxiang XuKejun 《高校应用数学学报(英文版)》1999,14(3):329-336
In this paper,a semlparametrie resresaion model in which errors are i. i. d random variables from an unknown density f( ) is considered. Based on Hall et al. (1995),a nonlinear wavelet estimation of f( ) without restrictions of continuity everywhere on f( ) is given,and the convergence rate of the estimators in L2 is obtained. 相似文献
5.
We propose to approximate the conditional density function of a random variable Y given a dependent random d-vector X by that of Y given θ^τX, where the unit vector θ is selected such that the average Kullback-Leibler discrepancy distance between the two conditional density functions obtains the minimum. Our approach is nonparametric as far as the estimation of the conditional density functions is concerned. We have shown that this nonparametric estimator is asymptotically adaptive to the unknown index θ in the sense that the first order asymptotic mean squared error of the estimator is the same as that when θ was known. The proposed method is illustrated using both simulated and real-data examples. 相似文献
6.
Chen Xiru 《数学年刊B辑(英文版)》2001,22(4):471-474
Consider,the linear regression modelyi = x'iβ ei, 1≤i≤n, n≥1, (1)where x1, x2,' are known Hvectors, P is the unknown pdimensional vector of regressioncoefficiellts, e13 e2,' is a seqdence of iid. random errors, and y1, y2t' are known obser-vations of the dependellt variable. Denote by F the common distribution of e1, e2,' t andwrite9. = {F: j:xar=0, 0< j:lxl'aF< oo}, 1 5 r 5 2.The Least Squares (LS) estimate of P isn rs)n = Z SJ'x.ui, Sn = Zxix:.. i=1 i=1Here we tacitly assum… 相似文献
7.
Jin-hong You Gemai Chen Min Chen Xue-lei JiangUniversity of Regina Regina Saskatchewan SS OA CanadaUniversity of Calgary Calgary Alberta TN N CanadaAcademy of Mathematics System Sciences Chinese Academy of Sciences Beijing China 《应用数学学报(英文版)》2003,19(3):363-370
Consider the partly linear regression model ,where yi's are responses, xi = (xi1, xi2,…,xip)' and ti ∈T are known and nonrandom design points, T is a compact set in the real line is an unknown parameter vector, g(·) is an unknown function and {Ei} isa linear process, i.e., random variables with zeromean and variance o2e. Drawing upon B-spline estimation of g(·) and least squares estimation of 0, we construct estimators of the autocovariances of {Ei}- The uniform strong convergence rate of these estimators to their true values is then established. These results not only are a compensation for those of [23], but also have some application in modeling error structure. When the errors {Ei} are an ARMA process, our result can be used to develop a consistent procedure for determining the order of the ARMA process and identifying the non-zero coefficients of the process. Moreover, our result can be used to construct the asymptotically efficient estimators for parameters in the ARMA error process. 相似文献
8.
Let {(ξni, ηni), 1 ≤ i ≤ n, n ≥ 1} be a triangular array of independent bivariate elliptical random vectors with the same distribution function as(S_1, ρ_n S_1 +(1-ρ_n~2S_2)~(1/2)), ρn∈(0, 1), where(S1, S2) is a bivariate spherical random vector. For the distribution function of radius (S_1~2+ S_2~2)~(1/2) belonging to the max-domain of attraction of the Weibull distribution, the limiting distribution of maximum of this triangular array is known as the convergence rate of ρn to 1 is given. In this paper,under the refinement of the rate of convergence of ρn to 1 and the second-order regular variation of the distributional tail of radius, precise second-order distributional expansions of the normalized maxima of bivariate elliptical triangular arrays are established. 相似文献
9.
This paper is concerned with the estimating problem of a semiparametric varying-coefficient partially linear errors-in-variables model Yi=Xτiβ+Zτiα(Ui)+εi , Wi=Xi+ξi,i=1, ··· , n. Due to measurement errors, the usual profile least square estimator of the parametric component, local polynomial estimator of the nonparametric component and profile least squares based estimator of the error variance are biased and inconsistent. By taking the measurement errors into account we propose a generalized profile least squares estimator for the parametric component and show it is consistent and asymptotically normal. Correspondingly, the consistent estimation of the nonparametric component and error variance are proposed as well. These results may be used to make asymptotically valid statistical inferences. Some simulation studies are conducted to illustrate the finite sample performance of these proposed estimations. 相似文献
10.
The authors consider the partially linear model relating a response Y to predictors(x, T) with a mean function x Tβ0+ g(T) when the x s are measured with an additive error. The estimators of parameter β0are derived by using the nearest neighbor-generalized randomly weighted least absolute deviation(LAD for short) method. The resulting estimator of the unknown vector β0is shown to be consistent and asymptotically normal. In addition, the results facilitate the construction of confidence regions and the hypothesis testing for the unknown parameters. Extensive simulations are reported, showing that the proposed method works well in practical settings. The proposed methods are also applied to a data set from the study of an AIDS clinical trial group. 相似文献
11.
§1.IntroductionConsiderthemodelgivenbyYi=xiτβ g(ti) εi,i=1,…,(1.1)wherexi=(xi1,…,xip)τ(p1)andti(ti∈[0,1])areknowndesignpoints,β=(β1,…,βp)τisanunknownparametervectorandgisanunknownfunction,andεiarei.i.d.randomerrorswithzero0andvarianceσ2.Themodeldefinedin(1.1)belongstotheclass… 相似文献
12.
考虑半参数回归模型$y_i=x_i\beta+g(t_i)+V_i$ $(1\le i\len)$, 其中$(x_i,t_i)$是已知的设计点, 斜率参数$\beta$是未知的,$g(\cdot)$是未知函数, 误差$V_i=\tsm^\infty_{j=-\infty}c_je_{i-j}$,$\tsm^\infty_{j=-\infty}|c_j|<\infty$并且$e_i$是负相关的随机变量.在适当的条件下, 我们研究了$\beta$与$g(\cdot)$小波估计量的强收敛速度.结果显示$g(\cdot)$的小波估计量达到最优收敛速度. 同时,对$\beta$小波估计量也作了模拟研究. 相似文献
13.
考虑回归模型:Y~((j))(x_(in),t_(in))=t_(in)β+g(x_(in))+σ_(in)e~((j))(x_(in)),1≤j≤m,1≤i≤n,其中σ_(in)~2=f(u_(in)),(x_(in),t_(in),u_(in))为固定非随机设计点列,β是未知待估参数,g(·)和f(·)是未知函数,误差{e~((j))(x_(in))}是均值为零的NA变量.给出基于g(·)和f(·)一类非参数估计的β的最小二乘估计和加权最小二乘估计,并在适当条件下得到了它们的强相合性. 相似文献
14.
THE RATES OF CONVERGENCE OF M-ESTIMATORS FOR PARTLY LINEAR MODELS IN DEPENDENT CASES 总被引:2,自引:0,他引:2
51.IntroductionAnembeddingofacoveringmaporaprincipalG-bundleT:E- Xintoavectorbundlep:V→XisamapH:E-VwhichmapsEhomeomorphicallyontoitsimageH(E)inVsuchthatpoH=7.In[6l,HansenprovedthatanyfinitecoveringmapoveraCW-complexcanbeembeddedintothetrivialrealm-planebundleifdimX5m 1.Embeddingfinite..covringmapsintoarbitraryvectorbundleswasconsideredbyDuval1andHuschl3].Suchanembeddingproblemwasalsodiscussedin[4,5,6,7,8,11,12]and[13].Inthisnote,thefirstpoilltistostudyingreaterdepththeembeddingproble… 相似文献
15.
本文研究一类二阶脉冲微分方程:■的正解存在性.其中,0<η<1,0<α<1,f:[0,1]×[0,∞)×R→[0,∞),I_i:[0,∞)×R→R,J_i:[0,∞)×R→R,(i=1,2,…,k)均为连续函数.本文所用方法是文献[5]推广的Krasnoselskii不动点定理,此定理为解决依赖于一阶导数的边值问题提供了理论依据.基于此定理,获得了问题正解存在性定理.特别地,我们获得此类问题的Green函数,使问题的解决更直观和简单. 相似文献
16.
本文研究了下列变系数混合效应模型: $y_{ij}=z_{ij}^{\tau}b_i+x_{ij}^{\tau}\beta(w_{ij}) +\xe_{ij},\;i=1,\cdots,m;\;j=1,\cdots,n_i$, 其中$b_i$为i.i.d.期望为$\xt$, 协方差阵为$\xs^2_bI_q$的随机效应向量, $\xe_{ij}$是i.i.d.期望为零, 具有有限方差的随机误差. 文中我们不仅给出了函数系数向量$\xb(\cdot)$的局部多项式估计, 同时给出了随机效应期望、方差和随机误差方差的估计, 并给出了这些估计量的渐进正态性和相合性, 研究结果表明了这些估计量的可靠性. 相似文献
17.
BOUNDARYVALUEPROBLEMSOFSINGULARLYPERTURBEDINTEGRO-DIFFERENTIALEQUATIONSZHOUQINDEMIAOSHUMEI(DepartmentofMathematics,JilinUnive... 相似文献
18.
Astrid Baumann 《Aequationes Mathematicae》2003,65(3):201-235
Summary. Let $\widehat{\widehat T}_n$ and $\overline U_n$ denote the modified Chebyshev polynomials defined by $\widehat{\widehat T}_n (x) = {T_{2n + 1} \left(\sqrt{x + 3 \over 4} \right) \over \sqrt{x + 3 \over 4}}, \quad \overline U_{n}(x) = U_{n} \left({x + 1 \over 2}\right) \qquad (n \in \mathbb{N}_{0},\ x \in \mathbb{R}).$ For all $n \in \mathbb{N}_{0}$ define $\widehat{\widehat T}_{-(n + 1)} = \widehat{\widehat T}_n$ and $\overline U_{-(n + 2)} = - \overline U_n$, furthermore $\overline U_{-1} = 0$. In this paper, summation formulae for sums of type $\sum\limits^{+\infty}_{k = -\infty} \mathbf a_{\mathbf k}(\nu; x)$ are given, where
$\bigl(\mathbf a_{\mathbf k}(\nu; x)\bigr)^{-1} = (-1)^k \cdot \Bigl( x \cdot \widehat{\widehat T}_{\left[k + 1 \over 2\right] - 1} (\nu) +\widehat{\widehat T}_{\left[k + 1 \over 2\right]}(\nu)\Bigr) \cdot \Bigl(x \cdot \overline U_{\left[k \over 2\right] - 1} (\nu) + \overline U_{\left[k \over 2\right]} (\nu)\Bigr)$ with real constants $ x, \nu $. The above sums will turn out to be telescope sums. They appear in connection with projective geometry. The directed euclidean measures of the line segments of a projective scale form a sequence of type $(\mathbf a_{\mathbf k} (\nu;x))_{k \in \mathbb{Z}}$ where $ \nu $ is the cross-ratio of the scale, and x is the ratio of two consecutive line segments once chosen. In case of hyperbolic $(\nu \in \mathbb{R} \setminus] - 3,1[)$ and parabolic $\nu = -3$ scales, the formula $\sum\limits^{+\infty}_{k = -\infty} \mathbf a_{\mathbf k} (\nu; x) = {\frac{1}{x - q_{{+}\atop(-)}}} - {\frac{1}{x - q_{{-}\atop(+)}}} \eqno (1)$ holds for $\nu > 1$ (resp. $\nu \leq - 3$), unless the scale is geometric, that is unless $x = q_+$ or
$x = q_-$. By $q_{\pm} = {-(\nu + 1) \pm \sqrt{(\nu - 1)(\nu + 3)} \over 2}$ we denote the quotient of the associated geometric sequence.
相似文献
19.
BAHADURASYMPTOTICEFFICIENCYINASEMIPARAMETRICREGRESSIONMODEL¥LIANGHUA;CHENGPINGAbstract:TheauthorSgiveMLEθ1MLofθ1inthemodelY=θ... 相似文献
20.
《复变函数与椭圆型方程》2012,57(2):95-110
Let $ k \in {\shadN} $ , $ w(x) = (1+x^2)^{1/2} $ , $ V^{\prime} _k = w^{k+1} {\cal D}^{\prime} _{L^1} = \{{ \,f \in {\cal S}^{\prime}{:}\; w^{-k-1}f \in {\cal D}^{\prime} _{L^1}}\} $ . For $ f \in V^{\prime} _k $ , let $ C_{\eta ,k\,}f = C_0(\xi \,f) + z^k C_0(\eta \,f/t^k)$ where $ \xi \in {\cal D} $ , $ 0 \leq \xi (x) \leq 1 $ $ \xi (x) = 1 $ in a neighborhood of the origin, $ \eta = 1 - \xi $ , and $ C_0g(z) = \langle g, \fraca {1}{(2i \pi (\cdot - z))} \rangle $ for $ g \in V^{\,\prime} _0 $ , z = x + iy , y p 0 . Using a decomposition of C 0 in terms of Poisson operators, we prove that $ C_{\eta ,k,y} {:}\; f \,\mapsto\, C_{\eta ,k\,}f(\cdot + iy) $ , y p 0 , is a continuous mapping from $ V^{\,\prime} _k $ into $ w^{k+2} {\cal D}_{L^1}$ , where $ {\cal D}_{L^1} = \{ \varphi \in C^\infty {:}\; D^\alpha \varphi \in L^1\ \forall \alpha \in {\shadN} \} $ . Also, it is shown that for $ f \in V^{\,\prime} _k $ , $ C_{\eta ,k\,}f $ admits the following boundary values in the topology of $ V^{\,\prime} _{k+1} : C^+_{\eta ,k\,}f = \lim _{y \to 0+} C_{\eta ,k\,}f(\cdot + iy) = (1/2) (\,f + i S_{\eta ,k\,}f\,); C^-_{\eta ,k\,}f = \lim _{y \to 0-} C_{\eta ,k\,} f(\cdot + iy)= (1/2) (-f + i S_{\eta ,k\,}f ) $ , where $ S_{\eta ,k} $ is the Hilbert transform of index k introduced in a previous article by the first named author. Additional results are established for distributions in subspaces $ G^{\,\prime} _{\eta ,k} = \{ \,f \in V^{\,\prime} _k {:}S_{\eta ,k\,}f \in V^{\,\prime} _k \} $ , $ k \in {\shadN} $ . Algebraic properties are given too, for products of operators C + , C m , S , for suitable indices and topologies. 相似文献