时间序列模型L_1-估计量的极限分布:平稳自回旧模型

时间序列模型的L_1-估计问题是非常重要的．这些估计量的许多性质都被研究过．然而,仍有一些基本问题有待解决．本文将给出平稳自回归模型L_1-估计量的极限分布．     

时间序列模型L1-估计量的极限分布:平稳自回归模型

时间序列模型的L1-估计问题是非常重要的.这些估计量的许多性质都被研究过.然而,仍有一些基本问题有待解决.本文将给出平稳自回归模型L1-估计量的极限分布.     

分数扩散模型的参数估计及其应用

研究分数扩散模型的参数估计及其应用问题.分数扩散模型是一类由分数Brownian运动驱动的随机微分方程.主要结果有:(1)利用二次变差方法给出模型中扩散系数的估计量,通过最小二乘法给出模型中漂移系数的估计量;(2)证明这些估计量的一致收敛性和渐近正态性;(3)利用MCMC方法对此估计量进行验证,并通过R软件将上述模型以及参数估计量应用到SHIBOR利率中进行实证研究.     

固定设计下半参数回归模型估计的相合性

对于固定设计下的半参数模型yi=x1β g(ti)┬ei=1，2……，n本文综合最小二乘法和一般的非参数权估计方法，定义了β，g的估计量-βn，-gn及误差方差口α^2=Ee^21的估计量-α^2n，并在适当条件下，证明了它们的强相合性与P(≥2)阶平均相合性．     

局部平稳短期利率扩散模型的半参数估计

主要研究局部平稳扩散模型的半参数估计．首先,基于局部常数拟合,利用局部加权最小二乘法得到了漂移参数函数的估计量．同时,通过Kolmogorov向前方程,得到了扩散函数的估计量．然后,分别讨论了所得估计量的相合性和渐近正态性．最后,通过模拟研究说明了估计量的有效性．     

具有两个方差分量的线性模型的比较

考虑线性模型(?)其中 X 为已知 n×p 矩阵,V 是已知或未知的 n 阶非负定阵,β=(β_1,…,β_p)′∈R~p 是参向量.记具有结构(1.1)的模型为 L=(Y,Xβ,V).设有两个模型 L_1(Y_1,X_1β,V_1),L_2=(Y_2,X_2β,V_2),当 V_1,V_2已知,Ehrenfeld定义了 L_1优于 L_2的概念,并证明了当 V_1,V_2非奇异时,L_1优于 L_2当且仅当 X′_1V_1~(-1)X_1-X′_2V_2~(-1)X_2≥0(非负定);Stepniak,Wang and Wu 继续研究了 V_1,V_2奇异的情形;Stepniak and Torgersen 又定义了当 V_1,V_2具有形式 σ~2V(σ~2未知,V 已知)时,L_1优于 L_2的概念;而且 Stepniak 证明了 L_1优于 L_2当且仅当 X′_1(V_1+X_1X′_1)-X_1-X′_2(V_2+X_2X′_2)-X_2≥0.但是,我们知道,在许多统计问题中,可观察的随机向量 Y 的协方差阵 V却有这样的形式 V=∑θ_iV_i,这里θ_i 为未知参数.事实上,在求方差分量的估计时,由均值-方差对应法导出的新模型其新的协方差阵往往不具有 σ~2V 这么简洁的形式(参见[5]).本文考虑的模型是 L_i=(Y_i,X_i,σ_1~2U_i+σ_2~2V_i),这里 U_i,V_i 均为已知非负定阵;σ_1~2,σ_2~2为未知参数.我们将给出 L_1优于 L_2(记为 L_1(?)L_2)的定义及判定准则。     

α-混合的部分线性EV模型的矩收敛性

本文研究误差为α-混合的部分线性EV模型的矩收敛性问题.利用小波估计和修正最小二乘法,给出了参数和非参数部分的小波估计量,获得了小波估计量的矩收敛速度,推广了现有的一些结论.     

数量特征敏感性问题随机化回答的改进模型

研究数量特征敏感性问题的调查方法,设计了一种数量特征敏感性问题的随机化回答改进模型,计算了改进模型的估计量及其方差,并对改进模型进行了分析,得出改进模型具有较好的精度．     

协变量缺失下部分线性模型的模型选择和模型平均

本文研究了协变量随机缺失下部分线性模型的模型选择和模型平均问题.首先利用逆概率加权方法得出了线性回归系数和非参数函数的估计,并在局部误设定框架下证明了线性回归系数估计量的渐近正态性.然后构造了兴趣参数的兴趣信息准则和频数模型平均估计量,并根据该模型平均估计量构造了一个覆盖真实参数的概率趋于预定水平的置信区间.模拟研究和实例分析分别说明了本方法的优越性和实用性.     

连续变量隐私问题的随机化回答模型

本文讨论连续变量隐私问题的抽样调查方法,建立了两个随机化回答模型,并在上述模型下给出总体均值的无偏估计、估计量的方差及方差的估计量     

Robust L1-norm non-parallel proximal support vector machine

In this paper, we propose a robust L1-norm non-parallel proximal support vector machine (L1-NPSVM), which aims at giving a robust performance for binary classification in contrast to GEPSVM, especially for the problem with outliers. There are three mainly properties of the proposed L1-NPSVM. Firstly, different from the traditional GEPSVM which solves two generalized eigenvalue problems, our L1-NPSVM solves a pair of L1-norm optimal problems by using a simple justifiable iterative technique. Secondly, by introducing the L1-norm, our L1-NPSVM is more robust to outliers than GEPSVM to a great extent. Thirdly, compared with GEPSVM, no parameters need to be regularized in our L1-NPSVM. The effectiveness of the proposed method is demonstrated by tests on a simple artificial example as well as on some UCI datasets, which shows the improvements of GEPSVM.     

ERROR ESTIMATES FOR SPARSE OPTIMAL CONTROL PROBLEMS BY PIECEWISE LINEAR FINITE ELEMENT APPROXIMATION

Optimization problems with L¹-control cost functional subject to an elliptic partial differential equation(PDE)are considered.However,different from the finite dimensiona l¹-regularization optimization,the resulting discretized L¹norm does not have a decoupled form when the standard piecewise linear finite element is employed to discretize the continuous problem.A common approach to overcome this difficulty is employing a nodal quadrature formula to approximately discretize the L¹-norm.In this paper,a new discretized scheme for the L¹-norm is presented.Compared to the new discretized scheme for L¹-norm with the nodal quadrature formula,the advantages of our new discretized scheme can be demonstrated in terms of the order of approximation.Moreover,finite element error estimates results for the primal problem with the new discretized scheme for the L¹-norm are provided,which confirms that this approximation scheme will not change the order of error estimates.To solve the new discretized problem,a symmetric Gauss-Seidel based majorized accelerated block coordinate descent(sGS-mABCD)method is introduced to solve it via its dual.The proposed sGS-mABCD algorithm is illustrated at two numerical examples.Numerical results not only confirm the finite element error estimates,but also show that our proposed algorithm is efficient.     

QFS-domains and quasicontinuous domains

In this paper, we show that(1) for each QFS-domain L, L is an ωQFS-domain iff L has a countable base for the Scott topology;(2) the Scott-continuous retracts of QFS-domains are QFSdomains;(3) for a quasicontinuous domain L, L is Lawson compact iff L is a finitely generated upper set and for any x1, x2 ∈ L and finite G1, G2■L with G1x1, G2x2, there is a finite subset F ■L such that ↑ x1 ∩↑ x2 ■↑ F ■↑ G1 ∩↑ G2;(4) L is a QFS-domain iff L is a quasicontinuous domain and given any finitely many pairs {(Fi, xi) : Fi is finite, xi ∈ L with Fi xi, 1 ≤ i ≤ n}, there is a quasi-finitely separating function δ on L such that Fi δ(xi) xi.     

Convergence of logarithmic means of multiple Walsh-Fourier series

Norlund logarithmic means of multiple Walsh-Fourier series acting from space Llnd-1 L ([0, 1)d) ,d≥1 into space weak-L1([0,1)d) are studied. The maximal Orlicz space such that the Norlund logarithmic means of multiple Walsh-Fourier series for the functions from this space converge in d-dimensional measure is found.     

Property ∏σ and tensor products of operator algebras

In this paper, we introduce Property ∏σ of operator algebras and prove that nest subalgebras and the finite-width CSL subalgebras of arbitrary von Neumann algebras have Property ∏σ.Finally, we show that the tensor product formula alg ML1-(×)algNL2 = algM-(×)N(L1 (×) L2) holds for any two finite-width CSLs L1 and L2 in arbitrary von Neumann algebras M and N, respectively.     

Stability and Convergence Analyses of the FDM Based on Some L-Type Formulae for Solving the Subdiffusion Equation

Some well-known L-type formulae, i.e., L1, L1-2, and L1-2-3 formulae, are usually employed to approximate the Caputo fractional derivative of order α ∈ (0, 1). In this paper, we aim to elaborate on the stability and convergence analyses of some finite difference methods (FDMs) for solving the subdiffusion equation, i.e., a diffusion equation which exploits the Caputo time-fractional derivative of order $α$. In fact, the FDMs considered here are based on the usual central difference scheme for the spatial derivative, and the Caputo derivative is approximated by using methods such as the L1, L1-2, and L1-2-3 formulae. Thanks to a specific type of the discrete version of the Gronwall inequality, we show that the FDMs are unconditionally stable in the maximum norm and also discrete $H^1$ norm. Then, we prove that the finite difference method which uses the L1, L1-2, and L1-2-3 formulae has the global order of convergence $2−α$, $3−α$, and 3, respectively. Finally, some numerical tests confirm the theoretical results. A brief conclusion finishes the paper.     

ON THE NUMBER OF INCREASING PATHS IN LABELED CYCLES AND STARS

A labeled graph is an ordered pair （G, L） consisting of a graph G and its labeling L ： V（G） → {1,2 ,n}, where n = ｜V（G）｜. An increasing nonconsecutive path in a labeled graph （G,L） is either a path （u1,u2 uk） （k ≥ 2） in G such that L（u,） ＋ 2 ≤ L（ui＋1） for all i = 1, 2, ..., k- 1 or a path of order 1. The total number of increasing nonconsecutive paths in （G, L） is denoted by d（G, L）. A labeling L is optimal if the labeling L produces the largest d（G, L）. In this paper, a method simpler than that in Zverovich （2004） to obtain the optimal labeling of path is given. The optimal labeling of other special graphs such as cycles and stars is obtained.     

对偶均值积分的Marcus-Lopes不等式

Milman曾提出过一个问题;在混合体积理论,是否存在Marcus-Lopes型和Bergstrom型不等式?即对R~n上任意凸体K与L且i=0,…,n-1,是否成立(W_i(K+L))/(W_i+1(K+L))≥(W_i(K))/(W_i+1(K))+(W_i(L))/(W_i+1(L))?这里W_i表示凸体的i次均值积分.当且仅当i=n-1或i=n-2时,这个问题是正确的,已被证明.作者考虑了一个对偶问题,证明了:若K与L是R~n上的星体,n-2≤i≤n-1且i∈R,则(W_i(K+L))/(W_i+1(K+L))≤(W_i(K))/(W_i+1(K))+(W_i(L))/(W_i+1(L))/(W_i+1(L))其中W_i表示星体的i次对偶均值积分.     

强奇异积分算子交换子的端点估计

本文考虑强奇异积分算子的交换子在Hardy型空间上的端点估计,建立了这类交换子从H¹（Rⁿ）到弱 L¹（Rⁿ）上的有界性及H¹（Rⁿ）的某个子空间到 L¹（Rⁿ）上的有界性结果.