门限自回归模型参数估计的渐近性质

本文在给定门限自回归模型阶数、门限及延迟参数的情况下,通过研究模型所构成Markov链的一些性质,得到了自回归系数最小二乘法估计的强相容性与渐近正态性。同时,还证明了白噪声方差的估计亦具有强相容性。     

残差自回归半参数模型的参数估计

研究了残差自回归半参数模型的参数估计,运用广义最小二乘法估计了参数部分．用随机模拟说明了运用广义最小二乘（GLSE）估计出的参数部分优于运用普通最小二乘法（OKSE）得到的估计．     

非线性E-V回归模型中参数估计的渐近性质

本文讨论了非线性E-V回归模型中参数的估计问题,构造了未知参数β0的最小二乘估计β和误差方差σ2的估计σ2,证明了β具有渐近正态性,同时也证明了σ2依概率收敛于σ2的速度可达到n-1/2.     

非线性E-V回归模型中参数估计的渐近性质

本文讨论了非线性E-V回归模型中参数的估计问题,构造了未知参数β0的最小二乘估计β和误差方差σ2的估计σ2,证明了β具有渐近正态性,同时也证明了σ2依概率收敛于σ2的速度可达到n-1/2.     

基于异方差的自回归预测模型的参数估计及其应用

从齐性方差的线性回归模型参数估计的最小二乘法出发,通过对统计资料的适当变换,利用加权最小二乘法,获得了异方差的线性自回归模型和四种异方差的非线性自回归模型的参数估计公式,并举例阐述了估计公式的应用.     

部分线性模型中估计的渐近正态性

考虑回归模型其中是未知函数，（ｘ＿ｉ，ｔ＿ｉ，ｕ＿ｉ）是固定非随机设计点列，β是待估参数，ｅ＿ｉ是随机误差。基于ｇ（·）及ｆ（·）的一类非参数估计（包括常见的核估计和近邻估计），我们构造了β的加权最小二乘估计，并证得了最小二乘估计和加权最小二乘估计的渐近正态性。     

多元线性回归最小二乘估计的一个非正常表现

通过一个实例揭示线性回归最小二乘估计的一个非正常性质,证明了:在误差方差无限时,在模型中人为地引入赘余参数有可能在相合性意义下改善原来的估计,并对这个现象的含义作了讨论。     

一类新的半参数回归模型中的相合估计

对一类新的半参数回归模型本文综合最小二乘和权函数估计方法，定义了β，ｇ的估计量β＿（ｍ，ｎ）和ｇ＿（ｍ，ｎ）（ｘ），在简洁合理的条件下，证明了它们具有强相合性与ｒ（＞２）阶平均相合性。     

重尾非线性自回归模型自加权M-估计的渐近分布

考虑非线性自回归模型xt=f(xt-1,…,xt-p,θ)+∈t,其中θ为q维未知参数,{∈t}为随机误差.在允许误差方差无穷的重尾条件下,构造θ的自加权M-估计,并证明了该估计的渐近正态性.最后通过数值模拟,在随机误差服从某些重尾分布的条件下,说明自加权M-估计比最小二乘和L1估计更有效.     

非参数异方差回归模型的局部多项式估计——基于农村居民消费与收入的实证分析

本文主要研究非参数异方差回归模型的局部多项式估计问题.首先利用局部线性逼近的技巧,得到了回归均值函数的局部极大似然估计.然后,考虑到回归方差函数的非负性,利用局部对数多项式拟合,得到了方差函数的局部多项式估计,保证了估计量的非负性,并证明了估计量的渐近性质.最后,通过对农村居民消费与收入的实证研究,说明了非参数异方差回归模型的局部多项式方法比普通最小二乘估计法的拟合效果更好,并且预测的精度更高.     

Local properties of algorithms for minimizing nonsmooth composite functions

This paper considers local convergence and rate of convergence results for algorithms for minimizing the composite functionF(x)=f(x)+h(c(x)) wheref andc are smooth buth(c) may be nonsmooth. Local convergence at a second order rate is established for the generalized Gauss—Newton method whenh is convex and globally Lipschitz and the minimizer is strongly unique. Local convergence at a second order rate is established for a generalized Newton method when the minimizer satisfies nondegeneracy, strict complementarity and second order sufficiency conditions. Assuming the minimizer satisfies these conditions, necessary and sufficient conditions for a superlinear rate of convergence for curvature approximating methods are established. Necessary and sufficient conditions for a two-step superlinear rate of convergence are also established when only reduced curvature information is available. All these local convergence and rate of convergence results are directly applicable to nonlinearing programming problems.This work was done while the author was a Research fellow at the Mathematical Sciences Research Centre, Australian National University.     

Convergence analysis of a monotone method for fourth-order semilinear elliptic boundary value problems

This work is concerned with the convergence of a monotone method for fourth-order semilinear elliptic boundary value problems. A comparison result for the rate of convergence is given. The global error is analyzed, and some sufficient conditions are formulated for guaranteeing a geometric rate of convergence.     

A convergence rate for approximate solutions of Fredholm integral equations of the first kind

We study the convergence rate for solving Fredholm integral equations of the first kind by using the well known collocation method. By constructing an approximate interpolation neural network, we deduce the convergence rate of the approximate solution by only using continuous functions as basis functions for the Fredholm integral equations of the first kind. This convergence rate is bounded in terms of a modulus of smoothness.     

Rates of Convergence of Ordinal Comparison for Dependent Discrete Event Dynamic Systems

Recent research has demonstrated that ordinal comparison, i.e., comparing relative orders of performance measures, converges much faster than the performance measures themselves do. Sometimes, the rate of convergence can be exponential. However, the actual rate is affected by the dependence among systems under consideration. In this paper, we investigate convergence rates of ordinal comparison for dependent discrete event dynamic systems. Although counterexamples show that positive dependence is not necessarily helpful for ordinal comparison, there does exist some dependence that increases the convergence rate of ordinal comparison. It is shown that positive quadrant dependence increases the convergence rate of ordinal comparison, while negative quadrant dependence decreases the rate. The results of this paper also show that the rate is maximized by using the scheme of common random numbers, a widely-used technique for variance reduction.     

DE-Sinc methods have almost the same convergence property as SE-Sinc methods even for a family of functions fitting the SE-Sinc methods

In this paper, the theoretical convergence rate of the trapezoidal rule combined with the double-exponential (DE) transformation is given for a class of functions for which the single-exponential (SE) transformation is suitable. It is well known that the DE transformation enables the rule to achieve a much higher rate of convergence than the SE transformation, and the convergence rate has been analyzed and justified theoretically under a proper assumption. Here, it should be emphasized that the assumption is more severe than the one for the SE transformation, and there actually exist some examples such that the trapezoidal rule with the SE transformation achieves its usual rate, whereas the rule with DE does not. Such cases have been observed numerically, but no theoretical analysis has been given thus far. This paper reveals the theoretical rate of convergence in such cases, and it turns out that the DE’s rate is almost the same as, but slightly lower than that of the SE. By using the analysis technique developed here, the theoretical convergence rate of the Sinc approximation with the DE transformation is also given for a class of functions for which the SE transformation is suitable. The result is quite similar to above; the convergence rate in the DE case is slightly lower than in the SE case. Numerical examples which support those two theoretical results are also given.     

Rates of convergence for continued fractions of irrational numbers

The dynamics of the Gauss Map suggests a way to compare the convergence to a real number ζ ε(0,l) of a continued fraction and the divergence of the orbit of ζ Of particular interest is the comparison of the rate of convergence to ζ of its simple continued fraction and the rate of divergence by the Gauss Map of the orbit of ζ for all irrational numbers in (0,l). We state and prove sharp inequalities for the convergence of the sequence of rational convergents of an irrational number ζ. We show that the product of the rate of convergence of the continued fraction of ζ and the rate of divergence by the Gauss Map of the orbit of ζ equals 1.     

Optimal Rates of Linear Convergence of Relaxed Alternating Projections and Generalized Douglas-Rachford Methods for Two Subspaces

We systematically study the optimal linear convergence rates for several relaxed alternating projection methods and the generalized Douglas-Rachford splitting methods for finding the projection on the intersection of two subspaces. Our analysis is based on a study on the linear convergence rates of the powers of matrices. We show that the optimal linear convergence rate of powers of matrices is attained if and only if all subdominant eigenvalues of the matrix are semisimple. For the convenience of computation, a nonlinear approach to the partially relaxed alternating projection method with at least the same optimal convergence rate is also provided. Numerical experiments validate our convergence analysis     

Constants in the estimates of the rate of convergence in von Neumann’s ergodic theorem with continuous time

Estimates for the rate of convergence in ergodic theorems are necessarily spectral. We find the equivalence constants relating the polynomial rate of convergence in von Neumann’s mean ergodic theorem with continuous time and the polynomial singularity at the origin of the spectral measure of the function averaged over the corresponding dynamical system. We also estimate the same rate of convergence with respect to the decrease rate of the correlation function. All results of this article have obvious exact analogs for the stochastic processes stationary in the wide sense.     

Tight global linear convergence rate bounds for Douglas–Rachford splitting

Recently, several authors have shown local and global convergence rate results for Douglas–Rachford splitting under strong monotonicity, Lipschitz continuity, and cocoercivity assumptions. Most of these focus on the convex optimization setting. In the more general monotone inclusion setting, Lions and Mercier showed a linear convergence rate bound under the assumption that one of the two operators is strongly monotone and Lipschitz continuous. We show that this bound is not tight, meaning that no problem from the considered class converges exactly with that rate. In this paper, we present tight global linear convergence rate bounds for that class of problems. We also provide tight linear convergence rate bounds under the assumptions that one of the operators is strongly monotone and cocoercive, and that one of the operators is strongly monotone and the other is cocoercive. All our linear convergence results are obtained by proving the stronger property that the Douglas–Rachford operator is contractive.     

Natural estimates of convergence rate in the central limit theorem

The concept of natural convergence rate estimates in the central limit theorem is proposed connecting convergence criteria and convergence rate. Supported by the Russian Foundation for Fundamental Research (grant No. 96-01-01920). Proceedings of the Seminar on Stability Problems for Stochastic Models. Hajdúszoboszló. Hungary. 1997, Part II.