量测误差为 ARMA 过程的随机逼近

为了求回归方程 h(x)=0的根 x~0,根据对回归函数 h(·)的量测,在 i 时刻对x~0的估计为 x_i,在 i+1时刻对回归函数在 x_i 处进行量测,但量测量 y_(i+1)带有误差ε_i:y_(i+1)=h(x_i)+ε_i,而误差是相关的,构成一个 ARMA 过程:ε_(n+1)+D_1ε_n+…+D_dε_(n-d+1)=ω_(n+1)(x_n,ω)+C_1ω_n(x_(n-1),ω)+…+C_rω_(n-r+1)(x_(n-r),ω),其中 ω_(i+1)(x_i,ω)是一个鞅差序列,熟知的定理讨论的是 d=0,r=0的特例,并要求 ω_(i+1)(x_i,ω)相互独立.本文给出一个随机逼近算法,并给出条件,当 n→∞时,x_n(?)x~0 a.s..这个结果对d=0,r=0的特例,和熟知的事实相比,不仅在噪声的性质上,而且对 h(·)及E‖ω_(n+1)(x,ω)‖~2的控制函数,y_(i+1)和 x_i 的维数差别等方面都减弱了条件.     

关于保凸Spline插值方法

§1.引言 本文讨论保凸插值方法和单调保凸插值问题.设a=x_0相似文献   

混合回归模型误差方差M估计的弱收敛性

考虑混合回归模型 y_i=x_i~Tβ+σε_i,(1)其中x_i~T=(y_(i-1),…,y_(i-p),z_(i1),…,z_(ik)),{ε_i}为i.i.d.残差序列,Eε1=0,Eε_1~2=1,而β=(β_1,…,β_p,β_(p+1),…,β_(p+k))~T与σ>0为未知参数,并且φ(B)=1-β_1B-…-β_pB~p=0的根全在单位圆外. 本文拟在文[1]的基础上定义模型(1)误差方差σ的M估计,并证明其弱收敛性. 设X(x)为某个可测函数,β为(1)中回归参数β的某个相容估计,称方程     

从函数的极大极小值近似恢复函数和它的极值点位置

1 问题的提法 已知一定义在[a,b]中上的函数f(x)在k个内点(x_i)_(i=1)~k处的极大和极小值(y_i)_(i=1)~k和两个端点值y_0,y_(k+1).其中 a=x_0相似文献   

分段线性规划算法的一个注记

求解分段线性规划问题inf S(x) s.t.Ax≤b 0≤x≤(?) (1)其中(?)=((?)_1,(?)_2,…,(?)_n)及 b=(b_1,b_2,…,b_m)~T 是已知向量,A 是已知的(m,n)矩阵,元素为 α_(ij)。目标函数 s(x)是分段线性函数。即对[0,(?)_j)(j=1,2,…,n)存在分划0=x_j~(0)相似文献   

加权回归模型及其在外延水文变量中的应用

§1.加权回归模型为了外延极大值水文变量,本文提出了一种权残差平方和目标函数,并用最小二乘法建立起加权回归模型。其模型为:y_4~*=A+Bx_4~*+8_4 (1)其中8_4～N(O,σ~2K_T~(-2)),当i=1,2,…,l时,K_T=K_2;当i=l+1,l+2,…,n时,K_T=K_1,(x_4~*,y_4~*)由原始变量(x_,y_4)变换而得,即把y_4从大到小排列。得y_4~*,对应的x_4记为x_4~*。加权回归模型的估计值为y~*=x+bx~*,目标函数为权残差平方和最小。即     

一个三次样条插值定理

C.Davis和W.J.Kammerer曾先后用不同的方法证明了如下定理: 设y_0,y_1,…,y_n为实数,满足y_0>y_1,y_1y_3,…,则存在唯一的一个n次多项式P_n(x)和一组点x_0,x_1,…,x_n使得P_n(x_i)=y_i(i=0,1,…,n),P′_n(x_i)=0(i=1,2,…,n-1),0=x_0相似文献   

用CV和GCV方法估计非参数回归函数

设非参数回归模型为y_i=f(x_4)+ε_i,i=1,…,n,f(x)是[0,1]上未知的非参数回归函数。f(x)的核估计具有一个光滑参数h,分别利用CV和GCV准则来选择光滑参数h,得到f(x)的优良的非参数估计。假设{ε_4}是i.i.d.的r.v.s.,在ε_4的4阶矩有限的条件下,所选择出来的核估计及相应的Stein估计是相合的。     

U-统计量的精致渐近性

设{X_n．n≥1}是一非退化的i．i．d．随机变量序列,U_n是以二维Borel可测对称函数h(x,y)为核函数的U-统计量．记U_n=2/(n(n-1))Σ_≤i≤j≤nh(X_i,X_j)．本文分别在核函数h(x,y)只有4/3阶矩或4/3+δ,0＜δ≤1的情况下,对非常广泛的一类权函数(x)与边界函数b(x)得到了如下关于U-统计量U_n的精致渐近性:不仅使得已有的结果成为我们的特况,还大大降低了其中的矩条件．     

一维问题的一种积分形式的流量重构算法

1引言我们考虑如下一维二阶椭圆边界值问题(-(β(x)p′)(x))′=f(x),x∈(a,b) p(a)=p(b)=0(1))其中β=β(x)是一恒正函数,且β∈H~1(a,b),f∈L~2(a,b)．事实上,在此条件下,我们可保证p∈H~2(a,b)(见[1],[2])．(1)之弱形式为：求p∈H_0~1(a,b)使得a(p,q)=(f,q),(?)q∈H_0~1(a,b),(2)其中a(p,q)=(?)_a~bβp′q′dx,(f,g)=(?)_a~bfqdx．给定(a,b)的一个分割α=x_0＜x_1＜…＜x_(n-1)＜x_n=b,令h=(?)(x_i-x_(i-1)),(?)_i表示通常相应于节点x_i的形状函数,即(?)_i是连续的分段线性函数且满足(?)_i(x_k)=δ_(ik),这里δ_(ik)=(?)i,k=0,1,…,n．又记V_h~0=span{(?)_1,(?)_2,…,(?)_(n-1)),取V_h~0作为p的逼近空间,则求解(1)的标准有限元格式为：求ph∈V_h~0使得     

The generalized Roper-Suffridge extension operator in Banach spaces (III)

Suppose that X is a complex Banach space with the norm ‖·‖ and n is a positive integer with dim X ⩾ n ⩾ 2. In this paper, we consider the generalized Roper-Suffridge extension operator $ \Phi _{n,\beta _2 ,\gamma _2 , \ldots ,\beta _{n + 1} ,\gamma _{n + 1} } (f) $ \Phi _{n,\beta _2 ,\gamma _2 , \ldots ,\beta _{n + 1} ,\gamma _{n + 1} } (f) on the domain $ \Omega _{p_1 ,p_2 , \ldots ,p_{n + 1} } $ \Omega _{p_1 ,p_2 , \ldots ,p_{n + 1} } defined by

ON ADMISSffilLITY OF VARIANCE COMPONENTS ESTIMATES

Suppose that there is a variance components model $$\[\left\{ {\begin{array}{*{20}{c}} {E\mathop Y\limits_{n \times 1} = \mathop X\limits_{n \times p} \mathop \beta \limits_{p \times 1} }\{DY = \sigma _2^2{V_1} + \sigma _2^2{V_2}} \end{array}} \right.\]$$ where $\[\beta \]$,$\[\sigma _1^2\]$ and $\[\sigma _2^2\]$ are all unknown, $\[X,V > 0\]$ and $\[{V_2} > 0\]$ are all known, $\[r(X) < n\]$. The author estimates simultaneously $\[(\sigma _1^2,\sigma _2^2)\]$. Estimators are restricted to the class $\[D = \{ d({A_1}{A_2}) = ({Y^''}{A_1}Y,{Y^''}{A_2}Y),{A_1} \ge 0,{A_2} \ge 0\} \]$. Suppose that the loss function is $\[L(d({A_1},{A_2}),(\sigma _1^2,\sigma _2^2)) = \frac{1}{{\sigma _1^4}}({Y^''}{A_1}Y - \sigma _1^2) + \frac{1}{{\sigma _2^4}}{({Y^''}{A_2}Y - \sigma _2^2)^2}\]$. This paper gives a necessary and sufficient condition for $\[d({A_1},{A_2})\]$ to be an equivariant D-asmissible estimator under the restriction $\[{V_1} = {V_2}\]$, and a sufficient condition and a necessary condition for $\[d({A_1},{A_2})\]$ to equivariant D-asmissible without the restriction.     

A Block Chaotic and Asynchronous Algorithm for Consistent Systems with Incomplete Data

In this paper, we generalize the paracontracting matrices to pseudocontracting matrices. The convergence of (parallel) iteration $$x_i=P_{ji,x_{i-1}}x_{i-1}$$ and $$x_{i+r_i}=α_{j_i}x_{i+r_i-1}+(1-α_{j_i})P_{j_i,x_i}x_i$$ where $P_{j_i,x_i}, j=1,\cdots,n$, are paracontracting and/or pseudocontracting matrices is analyzed. These iterations can also be applied to solve consistent systems with incomplete data.     

多元线性模型中的有偏估计

刘金山在1999年给出了多元线性模型参数分量βi和参数矩阵B的有偏估计β1=(X'X)-1YCi,i=1,…,m和B=(β1,β2,…,βm)以及βi的性质.本文从另一角度得到了同样的估计,证明了刘金山文中所给的两个检验是UMP检验,估计βi是βi的线性可容许估计,证明了B不是B的可容许估计,由此给出了两种改进估计.     

ON SMOOTHNESS AND DIFFERENTIABILITY OF FUNCTIONS

Let \(f(x)\) be a bounded real function on [-1,1],we define the modulus of continuity of f as \[\omega (f,\delta ) = \mathop {\sup }\limits_{x,y \in [ - 1,1],\left| {x - y} \right| \le \delta } \left| {f(x) - f(y)} \right|\] and the modulus of smoothness of f as \[{\omega _2}(f,\delta ) = \mathop {\sup }\limits_{x \pm h \in [ - 1,1],\left| h \right| \le \delta } \left| {f(x + h) + f(x - h) - 2f(x)} \right|\] Functions \(f(x)\), continuous on [-1,1] and \({\omega _2}(f,\delta ) = o(\delta )\) ,are called uniformly smooth functions. It is well known that there is a uniformly smooth functions whose derivative exisits on a null-set only. It would is of interest to discuss what condition should be added on the nonnegative function \(\varphi (\delta )\), \(\left( {0 \le \delta \le \frac{1}{2}} \right)\),in order that every bounded function f satisfying\[{\omega _2}(f,\delta ) = O(\varphi (\delta ))\] possess continous (or finite) derivative. the main result of this paper are the following two theorems. Theorem 1 let \(\varphi (\delta )\),\(\left( {0 \le \delta \le \frac{1}{2}} \right)\) ,be a nonnegative function, then, in order that every bounded function \(f(x)\) satisfying condition \[{\omega _2}(f,\delta ) = O(\varphi (\delta ))\] possess continous (or finite) derivative \(f'(x)\) on [-1,1],it is necessary and sufficient that the following condition hold \[\int_0^{\frac{1}{2}} {\frac{{\tilde \varphi (t)}}{t}} dt < \infty \] where \[\tilde \varphi (\delta ) = {\delta ^2}\mathop {\inf }\limits_{0 \le \eta \le \delta } \left\{ {{\eta ^{ - 2}}\mathop {\inf }\limits_{\eta \le \xi \le 1/2} \varphi (\xi )} \right\}\] Theorm 2 Let \(f(x)\) be a bounded function with \[\int_0^{\frac{1}{2}} {\frac{{{\omega _2}(f,t)}}{{{t^2}}}} dt < \infty \] then \(f'(x)\) is a continous function and \[{\omega _2}(f',\delta ) = O\left\{ {\int_0^\delta {\frac{{{\omega _2}(f,t)}}{{{t^2}}}} dt} \right\}\].     

两类广泛递归序列的动力特性

We investigate the dynamics of two extensive classes of recursive sequences:xn+1=c∑ k ∑xn-ioxn-i1…xn-i2j+f(xn-io,xn-i1,…,xn-i2k)j=0(i0,i1,…,i2j)∈A2j/c∑ k ∑xn-ioxn-i1…xn-i2j-1+c+f(xn-io,xn-i1,…,xn-i2k)j=1(i0,i1,…,i2j)∈A2j-1 and xn+1=c∑ k ∑xn-ioxn-i1…xn-i2j-1+c+f(xn-io,xn-i1,…,xn-i2k)j=1(i0,i1,…,i2j)∈A2j-1/c∑ k ∑xn-ioxn-i1…xn-i2j+f(xn-io,xn-i1,…,xn-i2k)j=0(i0,i1,…,i2j)∈A2j We prove that their unique positive equilibrium x = 1 is globally asymptotically stable.And a new access is presented to study the theory of recursive sequences.     

Global Behavior of Rational Sequences Involving Piecewise Power Function

In this paper, we are concerned with the global behavior ofthe solutions of the difference equation

ON THE CONTACT COHOMOLOGY OF ISOLATED HYPERSURFACE SINGULARITIES

The author defines, using jets, cohomology $H^p(\Lambda _{f,k-})$ for hypersurfaces, which are invariant under contact transformations. For isolated hypersurface singularities, it is proved that $H^0(\Lambda _{f,k-})=O_{U,0}/f^{k+1}O_{U,0},$ $H^p(\Lambda _{f,k-})=0,1\leq p \leq N-3 or p=N,$ $dimH^{N-2}(\Lambda _{f,k-})-dimH^{N-1}(\Lambda _{f,k-})=\[\left( {\begin{array}{*{20}{c}} k \ N \end{array}} \right)\dim {O_{U,0}}/(f,\frac{{\partial f}}{{\partial {x_1}}}, \cdots ,\frac{{\partial f}}{{\partial {x_N}}}){O_{U,0}}\] $ The algorithm of computation for H^{N-2} and H^{N-1} is given, and it is proved that $H^{N-1}=0$ when f is quasi-homogeneous.     

Inequalities for sine sums with more variables

A result of Vietoris states that if the real numbers \(a_1,\ldots ,a_n\) satisfy

$$\begin{aligned} \text{(*) } \qquad a_1\ge \frac{a_2}{2} \ge \cdots \ge \frac{a_n}{n}>0 \quad \text{ and } \quad a_{2k-1}\ge a_{2k} \quad (1\le k\le n/2), \end{aligned}$$

then, for \(x_1,\ldots ,x_m>0\) with \(x_1+\cdots +x_m <\pi \),

$$\begin{aligned} \begin{aligned} \text{(**) } \qquad \sum _{k=1}^n a_k \frac{\sin (k x_1) \cdots \sin (k x_m)}{k^m}>0. \end{aligned} \end{aligned}$$

We prove that \((**)\) (with “\(\ge \)” instead of “>”) holds under weaker conditions. It suffices to assume, instead of \((*)\), that

$$\begin{aligned} \sum _{k=1}^N a_k \frac{\sin (kt)}{k}>0 \quad (N=1,\ldots ,n; \, 0<t<\pi ), \end{aligned}$$

and, moreover, \((**)\) is valid for a larger region, namely, \(x_1,\ldots ,x_m\in (0,\pi )\).

     

Functions for parametrization of solutions of an equation in a free monoid

In this paper we introduce recursive functions

of the word variables , natural number variables and variables whose values are finite sequences of natural number variables. By means of these functions we give finite expressions for the family of solutions of the equation

where is an arbitrary word in the alphabet , in a free monoid.