具有外生变量部分线性自回归模型的样条估计

考虑自回归模型Yt=θTXt+g(Zt)+Et,t=1,…,n,其中Xt=(Y-1,…,Yt-d)T,Zt为实值外生随机变量,θ=(θ1,…,θd)T为待估参数向量,g为未知非参数光滑函数.基于多项式样条方法,在一定的条件下,给出了θ的估计的渐近正态性,得到了g的估计的收敛速度.模拟例子验证了所得的理论结果.     

方差无穷非线性自回归序列的自加权L_1估计

对具有无穷方差的非线性自回归序列x_t=φ(x_(t-1),x_(t-2),…,x_(t-p),θ) ε_t,E(ε_t~2)=∞,利用局部二次近似和连续函数空间C(R~q)上弱收敛随机过程最小点的渐近性质,证明了若存在δ≥1,使得E|ε_t|~δ<∞成立,则θ满足一定条件的自加权L_1估计θ_(L_1)是渐近正态估计,Wald检验统计量也具有通常的x~2分布,为模型的统计推断提供了理论基础.     

具有外生变量部分线性自回归模型的样条估计

考虑自回归模型Y_t=θ~TX_t g(Zt) ε_t,t=1,…,n,其中X_t=(Y_(t-1),…,Y_(t-d))~T,Z_t为实值外生随机变量,θ=(θ_1,…,θ_d)~T为待估参数向量,g为未知非参数光滑函数．基于多项式样条方法,在一定的条件下,给出了θ的估计的渐近正态性,得到了g的估计的收敛速度．模拟例子验证了所得的理论结果．     

非线性自回归模型的分位数推断

考虑非线性自回归模型x_t=f (x_(t-1),···, x_(t-p),θ)+ε_t,其中f (·)为R~p上的实值可测函数,θ为q维未知参数,{ε_t}为随机误差.构造了θ的分位数估计,并证明了该估计的渐近正态性,最后通过数值模拟,说明了该估计的稳健和有效性.     

重尾扰动下近非平稳自回归模型的Bootstrap推断

考虑近非平稳一阶自回归模型,Xt=θXt-1+ut,其中θn=1-γ/n,γ为一固定常数,{ut}为一重尾随机扰动项.对θn及其最小二乘估计(θ)n,采用bootstrap再抽样方法逼近(θ)n-θn的分布,并证明了该再抽样方法的渐近有效性.     

带有重尾扰动项的非线性自回归模型

讨论在非线性自回归模型中平稳解边际尾概率与扰动项尾概率之间的关系. 证明了在保证模型平稳遍历的条件下, 一维平稳解的边际分布具有重尾概率性质, 并且与扰动项重尾概率指标相同.     

一类随机加权和的渐近性及其应用

本文考虑随机加权和及其最大值尾概率的渐近性,其中增量{X_i,i≥1}为一列独立同分布的实值随机变量,权重{θ_i,i≥1}为另一列非负的随机变量,并且两列随机变量满足某种相依结构.在增量的共同分布F属于控制变换分布族的条件下,我们得到了随机加权和及其最大值尾概率的弱渐近等价估计.特别地,当F属于一致变换分布族时,得到了渐近等价估计.最后,我们将该结果应用于破产概率的渐近估计.     

重尾相依随机变量的随机加权和尾概率的渐近估计（英文）

令{X_k;k≥1}为一列实值随机变量,{θ_k;k≥1}为另一列与之独立的随机变量序列.假设{X_k;k≥1}为两两广义负象限相依且服从重尾分布,在{θ_k;k≥1}独立和相依条件下,本文得到了一些渐近估计.     

双边截断型分布族参数函数的渐近最优经验Bayes估计

本文考虑一维双边截断型分布族参数函数在平方损失下的经验 Bayes估计问题 .给定θ,X的条件分布为f (x|θ) =ω(θ1,θ2 ) h(x) I[θ1,θ2 ] (x) dx其中θ =(θ1,θ2 )T(x) =(t1(x) ,t2 (x) ) =(min(x1,… ,xm) ,max(x1,… ,xm) )是充分统计量 ,其边缘密度为 f (t) ,本文通过 f (t)的核估计构造出θ的函数的经验 Bayes估计 ,并证明在一定的条件下是渐近最优的 (a.0 .)     

ACD模型自加权LAD估计的渐近性质北大核心

针对高频数据建模中常用的自回归条件持续期(ACD)模型,在允许误差方差无穷的条件下,构造模型参数的自加权最小一乘(SLAD)估计,并证明了该估计的相合性和渐近正态性.数值模拟显示SLAD估计比拟极大似然估计和最小一乘估计更稳健,最后将其应用于青岛海尔和宝信软件这两只股票的价格持续期建模.     

Efficient estimation and variable selection for infinite variance autoregressive models

In this paper, a self-weighted composite quantile regression estimation procedure is developed to estimate unknown parameter in an infinite variance autoregressive (IVAR) model. The proposed estimator is asymptotically normal and more efficient than a single quantile regression estimator. At the same time, the adaptive least absolute shrinkage and selection operator (LASSO) for variable selection are also suggested. We show that the adaptive LASSO based on the self-weighted composite quantile regression enjoys the oracle properties. Simulation studies and a real data example are conducted to examine the performance of the proposed approaches.     

THE CONSTRUCTION OF DENUMERABLE q-PROCESSES IN RANDOM ENVIRONMENTS SATISFYING (F) OR (B)

This article is a continuation of [9]. Based on the discussion of random Kol-mogorov forward (backward) equations, for any given q-matrix in random environment,Q(θ) = (q(θ; x, y), x, y ∈ X), an infinite class of q-processes in random environments sat-isfying the random Kolmogorov forward (backward) equation is constructed. Moreover,under some conditions, all the q-processes in random environments satisfying the random Kolmogorov forward (backward) equation are constructed.     

Inference for Spatial Autoregressive Models with Infinite Variance Noises

A self-weighted quantile procedure is proposed to study the inference for a spatial unilateral autoregressive model with independent and identically distributed innovations belonging to the domain of attraction of a stable law with index of stability α, α ∈ (0, 2]. It is shown that when the model is stationary, the self-weighted quantile estimate of the parameter has a closed form and converges to a normal limiting distribution, which avoids the difficulty of Roknossadati and Zarepour (2010) in deriving their limiting distribution for an M-estimate. On the contrary, we show that when the model is not stationary, the proposed estimates have the same limiting distributions as those of Roknossadati and Zarepour. Furthermore, a Wald test statistic is proposed to consider the test for a linear restriction on the parameter, and it is shown that under a local alternative, the Wald statistic has a non-central chisquared distribution. Simulations and a real data example are also reported to assess the performance of the proposed method.     

Upper functions for positive random functionals. I. General setting and Gaussian random functions

In this paper we are interested in finding upper functions for a collection of real-valued random variables {Ψ(χ _θ ), θ ∈ Θ}. Here {χ _θ , θ ∈ Θ} is a family of continuous random mappings, Ψ is a given sub-additive positive functional and Θ is a totally bounded subset of a metric space. We seek a nonrandom function U: Θ → ?₊ such that sup _θ∈Θ{Ψ(χ _θ ) ? U(θ)}₊ is “small” with prescribed probability. We apply the results obtained in the general setting to the variety of problems related to Gaussian random functions and empirical processes.     

混合缺失机制下两总体差异的半经验似然置信区间

假定两个总体x与y均有数据缺失,它们的分布函数分别为F(·)与G_θ(·),其中F(·)未知,G_θ(·)的概率密度函数g_θ(·)形式已知,仅依赖于一些未知的参数,利用Fractional填补法填补缺失值,在一定的条件下证明了缺失数据下两总体差异指标的半经验似然比统计量的渐近分布为x_1~2,由此可构造两总体差异指标的经验似然置信区间.     

带分布时滞中立型微分方程正解的存在性

主要运用Banach压缩映像原理,研究了分布时滞中立型微分方程ddty(t)+∫abP(t,θ)y(t-θ)dθ+∫cdQ(t,θ)y(t-θ)dθ=0正解的存在性问题,获得了正解存在的充分条件,推广了相关文献的结果.     

NA误差下半参数回归模型估计的相合性

对半参数回归模型:Y(xin,tin)=tinb+g(xin)+e(xin),1≤j≤m,1≤i≤n,本文在NA相依样本下讨论了g的加权估计及b的最小二乘估计的强相合性与r(>2)阶平均相合性,使得文献犤2犦在独立样本下的相应结果得到推广     

关于Neyman-Pearson基本引理的几个注记

本文探讨了Neyman-Pearson基本引理.通过论证总体参数θ只有θ₀或θ₁两种可能时最优检验功效函数的唯一性,得到了两种假设T₁:θ=θ₀←→θ=θ₁和T₂:θ=θ₁←→θ=θ₀各自对应最优检验的两类错误概率可以互换的结论.     

幂平均不等式的最优值

设Mn[r](a)为a的r阶幂平均,0<α<θ<β,那么满足不等式[Mn[α](a)]1-λ.[Mn[β](a)]λ≤Mn[θ](a)的最大实数λ是λ≥{1+(β-θ)/[m(θ-α)]}-1.这里m=min{[2+(n-2)tβ]/[2+(n-2)tα],t∈R++};满足反向不等式的最小实数λ是λ=[β(θ-α)]/[θ(β-α)].本文的方法基于优势理论与解析技巧,对于建立不等式的最优化思想作了尽可能多的展示.作为应用,得到了一些涉及和、积分与矩阵的新不等式(含Hardy不等式的推广与加强).     

二维Qausi-Geostrophic方程的几何约束与非爆炸性

研究二维Qausi-Geostrophic(QG)方程水平集的动态演化过程.在"涡线"上"涡度"的最大值与全局"涡度"最大值可比的假设下,得到若"涡线"的长度L(t)被O(ln ln ||▽~⊥θ(·,t)||_∞)控制(被O(ln ln ||▽~⊥θ(·,t)||_∞)控制),曲率最大值K(t)与L(t)的乘积被O(ln ln ln ||▽~⊥θ(·,t)||_∞)控制(被O(ln ln ||▽~⊥θ(·,t)||_∞)控制),则||▽~⊥θ(·,t)||_∞)的增长阶数最多可达四重指数阶,这样二维QG方程在有限时间内无爆炸发生.