相依样本下非参数回归函数估计的强收敛速度

设(X,Y),(X_1,Y_1,),…,(X_n,Y_n)是一个平稳、φ—混合过程((X,Y)∈R~d×R,E|Y|~(s δ)<∞,s≥2,δ>0),用m(x)记E{Y|X=x},本文讨论了m(x)的如下估计m_n(x)的强收敛速度:     

非参数回归核估计的条件平均绝对误差的指数收敛速度

设(X,Y)、(X,Y_1),…,(X_n,Y_n)是取值于 R~d×R~1的 iid。随机向量,E|Y|<∞,在本文中将一直采用下面的记号:Z_n={(X_i,Y_i),i=1,…,n}—(X,Y)的已知样本。X~n={X_1,…,X_n}。Q——X 的概率分布测度。m(x)=E(Y|X=x)——Y 对 X 的回归函数。现设有了 Z_(?)并指定了 R~d 中的一个点 x,要依据它们对 m(x)作出估计。这就是一般的非参数回归问题。核估计法就是先选定 R~d 上定义的非负函数 K(x)作为核函数,那么可给出 m(x)的一个核估计     

一种非参数回归估计量的条件 平均绝对误差的指数收敛速度

<正> 设(X,Y),(X_1,Y_1),…,(X_n,Y_n)是取值于R~d×R~1的独立同分布随机变量,E|Y|<∞.以m(x)=E(Y|X=x)记Y对X的回归函数,Q记X的概率分布测度,Z_n记{(X_i,Y_i),i=1,…,n},它是(X,Y)的已知观测值.一般的非参数回归估计问题,就是对指定的x∈R~d,利用Z_n对m(x)进行估计.设θ=θ(x,Z_n)是这样一个估     

非参数回归函数的随机窗宽核估计的重对数律

设(X,Y),(X_1,Y_1),(X_2,Y_2),…为 i.i.d.二维随机变量序列,具有联合分布F(x,y)及密度 f(x,y).X 的边际分布和密度分别记为 F_X(x)和 f_X(x).记 m(x)=E{Y|X=x)}为 Y 对 X 的回归函数.为估计 m(x),Nadaraya 和 watson 独立地引进了如下形式的核估计     

可加模型低维分量估计平均偏差的指数界

设(X,Z,y),(X1,Z1,Y1),…,(Xn,Zn,Yn)为取值于Rp×Rq×R中的I.I.d.随机向量,E|Y|<∞,Y关于(X,Z)的回归函数m(x,z)△E(Y|(X,Z)＝(x,z))具有可加结构:m(x,z)＝m1(x)+m2(z).为估计可加分量,采用Linton&Nielsen(1995)提出的直接估计法,给出了可加分量的最近邻估计和核估计.在较弱的条件下,建立了可加分量最近邻估计和核估计的平均偏差的指数界.     

非参数回归函数核估计的收敛速度

<正> §1.引言及记号设(Y,X),(Y_1,X_1),…,(Y_n,X_n)为 iid.(1+d)维随机向量,E(|Y|)<∞,m(x)=E(Y|X=x)为回归函数.Watson,Nadaraya 首先提出的基于样本(Y_1,X_1),…,(Y_n,X_n)的 m(x)的核估计为     

非参数回归函数之改良基于分割的估计的强相合性（英）

设(X,Y),(X1,Y1),…,(XnYn)为取值于 Rd× R的 i．i．d．随机变量,E(|Y|) ＜∞．设mn(x)为回归函数m(x)=E(|Y|X=x)基于分割的估计,本文在对mn(x)进行改良的条件下得到改良的基于分割的强相合估计．     

回归函数的相合随机窗宽核估计

令(X,Y)为取值于 R~d×R 的随机向量,(X_1,Y_1),……,(X_n,Y_n)为抽自(X,Y)的分布的 iid.样本,m(x)(?)E(Y|X=x)称为 Y 对 X 的回归函数.1964年,Watson和 Nadaraya 首先提出用     

回归函数改良核估计的相合性

一、引言及若干引理设(X_1,Y_1),…,(X_n,Y_n)为(X,Y)的前 n 个样本,(X,Y)为 R~d×R 上的随机向量,μ为 X 的概率分布,回归函数 m(x)=E(Y|X=x)的核估计为     

关于回归函数核估计的渐近正态性

令(X,Y)是具有联合密度f(x,y)的二元随机变量。如果EY有限,则称m(x)=E(Y|X=x)为Y关于X的迴归函数.假设(X_1,Y_1),…,(X_n,Y_n)是来自二元总体(X,Y)的一个随机样本,那么迴归函数的核估计定义作其中K是一元密度函数,{h_n}是一列收敛于0的正数.在Y有界且nh_n~2→∞的条件下,证明了(nh_n)~(1/2)(m_n(x)-Em_n(x))依分布收     

Strong Convergence of the Kernel Estimates of Nonparametric Regression Functions

Let (X, Y), (X_1, Y_1),\cdots, (X_n, Y_n) be i. i. d. random vectors taking values in R_d\times R with E(|Y|)<\infinity, To estimate the regression function m(x)=E(Y|X=x), we use the kernel estimate $m_n(x)=[\sum\limits_{i = 1}^n {K(\frac{{{X_i} - x}}{{{h_n}}}){Y_i}/} \sum\limits_{i = 1}^n {K(\frac{{{X_j} - x}}{{{h_n}}})} \]$ where K(x) is a kernel function and h_n a window width. In this paper, we establish the strong consistency of m_n(x) when E(|Y|^p)<\infinity for some p>l or E{exp(t|Y|^\lambda)}<\infinity for some \lambda>0 and t>0. It is remakable that other conditions imposed here are independent of the distribution of (X, Y).     

Almost Sure Convergence of Nonparametric Begression Estimates

Let (X, Y), (X_i,Y_i), i=1,\cdots, n, be iid.R^d*R^1-ralued vandom vectors with E(|Y|) <\infinity and m(x) =E(Y|X=x) be the regression function. Select the weight functions W_ni(x) =W_ni(x; X_1,\cdots, X_n), and use m_n(x) =[\sum\limits_{i = 1}^n {{W_{ni}}(x){Y_i}} \] an estimator of m(x). This paper shows that [\mathop {\lim }\limits_{n \to \infty } \]m_n(X) =m(X), a. s., under weaker conditions.     

Necessary Conditions of L_1-Convergenoe of Kernel Regression Estimators

Let (X_1,Y_1),\cdots,(X_n,Y_n) be iid. and R^d *R-valued samples of (X,Y). The kernel estimator of the regression function m(x)\triangleq E(Y|X=x) (if it exists), with kernel K, is denoted by $\[{m_n}(x) = \sum\limits_{i = 1}^n {{Y_i}K(\frac{{{X_i} - x}}{{{h_n}}})/\sum\limits_{j = 1}^n {K(\frac{{{X_j} - x}}{{{h_n}}})} } \]$ Many authors discussed the convergence of m_n(x) in various senses, under the conditions h_n\rightarrow 0 and nh_u^d\rightarrow \infinity asn\rightarrow \infinity. Are these conditions necessary? This paper gives an affirmative answer to this bprolemuithe case of L_1-conversence, when K satisfies (1.3) and E(|Y|log^+|Y|)<\infinity.     

相依样本下非参数回归函数递推型核估计的渐近分布

设（Ｘｉ，Ｙｉ）１≤ｉ≤ｎ为来自二元总体（Ｘ，Ｙ）的平稳，φ－混合样本，记ｍ（ｘ）△Ｅ（Ｙ│Ｘ＝ｘ），ｍ（ｘ）的一种递推型核估计为ｍｎ（ｘ）＝ｎ∑ｉ＝１ｈｉ＾－１Ｙｉｋ（（ｘ－Ｘｉ）／ｈｉ）／ｎ∑ｊ＝１ｈ＾－１ｊｋ（ｘ－Ｘｊ）／ｈｊ）。本文在一定的条件下证明了（ｎ／（ｎ∑ｊ＝１ｈ＾－１ｊ）＾１／２）（ｍｎ（ｘ１）－ｍ（ｘ１），ｍｎ（ｘ２）－ｍ（ｘ２），．．．ｍｎ（ｘｒ０）－ｍ（ｘｒ０））′依分布收     

A Simple Goodness-of-fit Test for Linear Models Under a Random Design Assumption

Let (X,Y) denote a random vector with decomposition Y = f(X) + where f(x) = E[Y ¦ X = x] is the regression of Y on X. In this paper we propose a test for the hypothesis that f is a linear combination of given linearly independent regression functions g1,..,gd. The test is based on an estimator of the minimal L2-distance between f and the subspace spanned by the regression functions. More precisely, the method is based on the estimation of certain integrals of the regression function and therefore does not require an explicit estimation of the regression. For this reason the test proposed in this paper does not depend on the subjective choice of a smoothing parameter. Differences between the problem of regression diagnostics in the nonrandom and random design case are also discussed.     

关于Thue方程的解数

设ｍ是正整数，ｆ（Ｘ，Ｙ）＝ａ０Ｘｎ＋ａ１Ｘ（ｎ－１）Ｙ＋．．．＋ａｎＹｎ∈Ｚ［Ｘ，Ｙ］是Ｑ上不可约化的叫ｎ（ｎ≥３）次齐次多项式。本文证明了：当ｇｃｄ（ｍ，ａ０）＝１，ｎ≥４００且ｍ≥１０（３５）时，方程｜ｆ（ｘ，ｙ）｜＝ｍ，ｘ，ｙ∈ｚ，ｇｃｄ（ｘ，ｙ）＝１，至多有６ｎｖ（ｍ）组解（ｘ，ｙ），其中ｖ（ｍ）是同余式Ｆ（ｚ）＝ｆ（ｚ，１）≡０（ｍｏｄｍ）的解数。特别是当ｇｃｄ（ｍ，ＤＦ）＝１时，该方程至多有６ｎ（ω（ｍ）＋１）组解（ｘ，ｙ），其中ＤＦ是多项式Ｆ的判别式，ω（ｍ）是ｍ的不同素因数的个数．     

加权经验累计分位函数的弱收敛

记(X,Y)为二元随机变量,F(x)为X的边缘分布函数.定义Y关于X的分位回归函数为h(u)=E(Y|F(X)=u),记S(u)=∫u0J(t)h(t)dt为加权累计分位回归函数,其中J(@)为权函数.本文讨论了S(u)的经验版本的弱收敛性质.     

ASYMPTOTICALLY ISOMETRIC COPIES OF lp (1≤p＜∞) AND c0 IN BANACH SPACES

Let X be a Banach space. If there exists a quotient space of X which is asymptotically isometric to l1, then X contains complemented asymptotically isometric copies of l1. Every infinite dimensional closed subspace of l1 contains a complemented subspace of l1 which is asymptotically isometric to l1. Let X be a separable Banach space such that X* contains asymptotically isometric copies of lp (1 ＜ p ＜∞). Then there exists a quotient space of X which is asymptotically isometric to lq (1/p 1/q=1). Complementedasymptotically isometric copies of c0 in K(X, Y) and W(X, Y) are discussed. Let X be a Gelfand-Phillips space. If X contains asymptotically isometric copies of c0, it has to contain complemented asymptotically isometric copies of c0.     

稳定性理論中第一临界情形的微分方程与微分差分方程的等价性問題

<正> §1.問題与方法.在[1]中提出了等价性問題,并对于一般n的情形作了系統的研究.本文是处理在第一临界情形下的微分方程与微分差分方程的等价性問題. 問題是研究微分方程組