部分线性函数多项式模型的联合探测北大核心CSCD

本文提出了一个新的部分线性函数多项式回归模型,该模型中响应变量依赖于一个p阶函数多项式和一些非函数型数据的协变量．函数多项式模型、函数线性模型和部分函数线性模型是该模型的特殊情形．本文提出了一个模型探测方法,它能同时探测部分线性函数多项式回归模型中哪些阶是重要的以及哪些非函数型变量是重要的．提出的方法能相合地识别真实的模型并有好的预测表现．数值模拟能清晰地证实我们的理论结果．     

通用回归分析模型

建立了通用的回归分析模型,它包含了直线、指数线、抛物线、双曲线等众多模型形式,因而具有广泛代表性.分析了通用模型中的常数项对模型曲线形状的影响.结合实例用计算机程序给出了相关参数的数值解法.通用模型与数值解法相结合,突破了回归分析中预先确定简单函数曲线误差大和复杂曲线求解难度大之间的矛盾问题,大幅度提高了计算精度和速度.     

利用平行数据模型研究中国20个省市消费、收入之间的关系

以中国20个省市的21年的城镇居民人均年生活费支出，城镇居民人均年可支配收入的平行数据为样本，建立消费函数模型．利用固定效应模型设定的假设检验方法，检验出中国20个省市之间不能建立统一的消费函数模型，将数据按消费倾向大分成三组分别建立消费函数模型．再利用LL平行数据单位根检验方法对三组消费、收入的平行数据进行单位根俭验，结果表明已建立的回归模型中不存在伪回归现象．     

分布估计算法中适应度函数模型检验及应用

在一种基于马尔可夫网络的分布估计算法中,利用解的适应度函数模型表示变量的概率分布.通过相关系数检验建立了一种衡量适应度函数模型有效性的方法,同时该方法也可以用来确定初始种群规模大小.在二维相关变量结构下,将该方法应用于一种多变量生物动态模型的优化问题,结论表明通过该方法可以选择适当的种群规模,保证适应度函数模型的有效性,并提高算法的优化性能.     

多类型复发事件下的一类均值加速回归模型

在复发事件的统计分析中,事件的平均发生个数可能比其强度函数或危险率函数更具可解释性.为了评价协变量对复发事件的影响,许多学者考虑了复发事件的边际比例均值模型.然而,在许多实际应用中,协变量对复发事件不仅具有均值比例效应,而且还可能会加速或减缓复发事件的发生,即协变量对复发事件均值过程具有时间尺度效应.在多类型复发事件数据框架下,考虑一类广泛的加速均值模型.利用估计方程方法,获得了该模型中未知参数的估计,并且建立了所给估计的渐近性质.进一步通过模拟研究证实所提方法的优良表现.     

稳健的模糊回归区间分析

由于传统的回归分析易受到异常值的影响。针对输入变量为实数,输出变量和输入参数为模糊数的情况,给出了一种稳健的模糊回归区间预测模型和算法。该模型基于输出变量的隶属度函数为目标函数,以估计的区间为约束条件。给出的算法具有较强的稳健性,利用该算法估计的区间几乎不受异常值的影响。最后通过一个数值算例,与其他模型算法对比分析,验证了该模型和算法的有效性和稳健性。     

基于一类重尾分布的函数型线性回归模型的稳健性估计

假定随机误差分布来自具有重尾特征的scale mixtures of normal分布族,运用贝叶斯方法研究了函数型线性回归模型的稳健性估计,其中模型的响应变量为标量,解释变量为函数型变量.数值模拟结果表明:当响应变量的观测数据存在离群值时,建立的方法得到的模型参数的估计,要优于正态分布假定下的模型参数的估计.     

具有测量误差的非线性模型的Bayes估计

测量中大量的函数模型都是非线性回归模型．当回归变量含有一定的测量误差时，我们得到非线性测量误差模型．本讨论了这种模型中未知参数具有正态先验分布时的参数Bayes估计方法，并对这种估计进行了影响分析，证明了删除模型与均值漂移模型中参数的Bayes估计相同，利用Cook统计量给出了删除模型下参数的Bayes估计的影响度量．     

广义部分线性单指标模型的最优模型平均方法

本文研究广义部分线性单指标模型(generalized partially linear single-index models, GPLSIMs)的模型平均问题.在实际应用中, GPLSIMs由于其灵活性和易解释性受到广泛关注.然而, GPLSIMs在应用中存在两类不确定性:变量的不确定性和单指标连接函数光滑度的不确定性.为了解决该不确定性问题,本文提出一种GPLSIMs的最优模型平均方法,该方法通过最大交叉验证准则得到数据驱动的权重.在模型误设定假设和发散模型空间的框架下,本文证明在最小化Kullback-Leibler (KL)损失准则下,所提出的模型平均估计渐近最优.同时,当候选模型集中存在伪真模型时,本文证明基于交叉验证准则得到的权重渐近地集中在伪真模型上.此外,基于提出的模型平均方法,本文为GPLSIMs构建了一种变量重要性度量,并证明该度量可以渐近识别所有真实模型中的变量.模拟研究和两个实际数据分析均展示了本文提出的方法相对于几种现有方法的优势.     

基于局部多项式回归的模型校准抽样估计研究

在抽样估计中,当研究变量与辅助变量之间呈非线性关系时,传统的校准估计方法效果较差,基于非参数回归方法的模型校准估计量则可以很好地解决这一问题。首先,建立描述研究变量和辅助变量之间关系的超总体回归模型,使用非参数中的局部多项式方法得出模型参数的拟合值,并结合校准估计得出局部多项式模型校准估计量,同时给出其方差和方差估计量公式,证明了该估计量具有渐近无偏性、一致性和渐近正态性等优良的统计性质。然后,使用仿真模拟的方法证明在研究变量与研究变量之间呈非线性关系时,该估计量有良好的估计效果。最后,对该估计量在我国政府统计中的应用进行简单的介绍。     

动态投入产出灰色最优控制模型

根据灰色系统理论,建立了动态投入产出问题的灰色最优控制模型.利用灰集合理论,把灰色最优控制问题转化为以隶属度为目标函数的(非灰色的)非线性规划问题,从而可利用非线性规划的方法求解这个灰色最优控制问题.     

应用Logistic模型对建三江农场水稻生产潜力的分析

本文首先建立了建三江农场水稻单产变化的 Logistic模型 ,通过对模型的微分方程形式的离散 ,得一二次函数 ,进行回归分析 ,得出模型中参数的估计值 .对模型进行分析得出现阶段水稻单产的最大潜力和水稻增产速度最快的时间段 ,对于指导水稻生产 ,优化农业生产结构 ,都具有很深的现实意义     

Convex composite non–Lipschitz programming

In this paper necessary, and sufficient optimality conditions are established without Lipschitz continuity for convex composite continuous optimization model problems subject to inequality constraints. Necessary conditions for the special case of the optimization model involving max-min constraints, which frequently arise in many engineering applications, are also given. Optimality conditions in the presence of Lipschitz continuity are routinely obtained using chain rule formulas of the Clarke generalized Jacobian which is a bounded set of matrices. However, the lack of derivative of a continuous map in the absence of Lipschitz continuity is often replaced by a locally unbounded generalized Jacobian map for which the standard form of the chain rule formulas fails to hold. In this paper we overcome this situation by constructing approximate Jacobians for the convex composite function involved in the model problem using ε-perturbations of the subdifferential of the convex function and the flexible generalized calculus of unbounded approximate Jacobians. Examples are discussed to illustrate the nature of the optimality conditions. Received: February 2001 / Accepted: September 2001?Published online February 14, 2002     

Teachers’ beliefs towards teaching calculus

This paper focuses on the belief systems towards teaching calculus of 29 upper secondary mathematics teachers. Firstly, we discuss different educational trends in teaching calculus, and the theoretical approach based on the construct of belief systems. Afterwards we describe the method of our study used to analyse the belief systems of the calculus teachers. Referring to findings of our research, we firstly focus on central and peripheral beliefs and thus discuss the structure of the teachers’ belief systems. Further, we compare the teachers’ belief systems to established educational trends of teaching calculus. Finally, we conclude the paper by reflecting on our main findings and by discussing possible directions of further research.     

First order optimality conditions in vector optimization involving stable functions

We study a nonsmooth vector optimization problem with an arbitrary feasible set or a feasible set defined by a generalized inequality constraint and an equality constraint. We assume that the involved functions are nondifferentiable. First, we provide some calculus rules for the contingent derivative in which the stability (a local Lipschitz property at a point) of the functions plays a crucial role. Second, another calculus rules are established for steady functions. Third, necessary optimality conditions are stated using tangent cones to the feasible set and the contingent derivative of the objective function. Finally, some necessary and sufficient conditions are presented through Lagrange multiplier rules.     

NA样本下半参数EV模型小波估计的渐近性简

In this paper, we research the semiparametric EV model under NA samples. Some estimators of the parameter, nonparameter and the variance function are established by the wavelet smoothing method. Under some general conditions, the strong consistency and the asymptotic normality of wavelet estimators are studied.     

A combination of time-scale calculus and a cross-validation technique used in fitting and evaluating fractional models

In this work, we demonstrate how fractional calculus and time-scale calculus can be combined beautifully to solve and fit a modeling problem. In addition, a cross-validation technique is used to evaluate the fitted model. The specific application that we consider is the one-compartment model. The one-compartment model is a first-order differential equation that describes drug concentration over time. It turns out that approximating the solution by using a fractional model allows us to get more accurate results for model fitting. To quantitatively verify this insight, we compare between a first-order model and anα-order fractional model using real data for drug concentration. Then the mean squared error and a cross-validation method are used to determine the model that provides the best fit and predictions for unseen data.     

由Dirichlet函数引出的微积分中若干反例

在本文中，我们主要利用Ｄｉｒｉｃｈｌｅｔ函数的构造特点给出微积分中的若干反例．     

Analysis of projectile motion in view of fractional calculus

The projectile motion is examined by means of the fractional calculus. The fractional differential equations of the projectile motion are introduced by generalizing Newton’s second law and Caputo’s fractional derivative is considered to use the physical initial conditions. In the absence of air resistance it is found that at certain conditions, the range and the maximum height of the projectile obtained by using the fractional calculus give the same results of the classical calculus. It is also found that, unlike the classical projectile motion, the launching angle that maximizes the horizontal range is a function of the arbitrary order of the fractional derivative α. Moreover, in a resistant medium, the parametric equations are expressed in terms of Mittag-Leffler function and the results agree with those of the classical projectile as α → 2. Moreover, the trajectories of the projectile are discussed in graphs and compared with those of the classical calculus. In order to explore the validity of modelling the projectile motion by the fractional approach, we compared our results with the experimental data of mortar.     

Non-Gaussian Malliavin calculus on real Lie algebras

The non-commutative Malliavin calculus on the Heisenberg-Weyl algebra is extended to the affine algebra. A differential calculus and a non-commutative integration by parts are established. As an application we obtain sufficient conditions for the smoothness of Wigner-type laws of non-commutative random variables with gamma or continuous binomial marginals.