固定效应部分线性单指数面板模型的快速有效估计及应用

将最小二乘支持向量机(LSSVM)和二次推断函数法(QIF)相结合,构造了个体内具有相关结构的固定效应部分线性单指数面板模型的新估计方法;在一定的正则条件下,证明了参数估计量的渐近正态性,导出了非参数估计量的收敛速度;Monte Carlo模拟了所述方法在各种相关结构下的有限样本表现,并与惩罚二次推断函数(PQIF)法进行了比较;将估计技术应用于分析我国人口结构与居民消费率的关系.研究发现,该方法改善了估计量的有效性,应用效果良好,程序运行速度快,适合经济变量间的线性和非线性关系研究以及大数据分析.     

固定效应部分线性可加动态面板模型的惩罚二次推断函数估计北大核心CSCD

工具变量法是估计动态面板模型的常用方法,但该方法并没有充分利用现有矩条件,导致所得估计有效性不足.为此,本文首先采用变量变换法消除模型的内生性,再用惩罚二次推断函数法推导出个体内具有一阶自相关结构的固定效应部分线性可加动态面板模型中未知参数和函数的估计;进一步,证明了所得估计量的一致性和渐近正态性,同时还用Monte Carlo模拟实验比较了该方法和半参数GMM法在有限样本下的表现;最后将所述方法应用于实际数据分析中.     

固定效应部分线性可加动态面板模型的惩罚二次推断函数估计

工具变量法是估计动态面板模型的常用方法,但该方法并没有充分利用现有矩条件,导致所得估计有效性不足.为此,本文首先采用变量变换法消除模型的内生性,再用惩罚二次推断函数法推导出个体内具有一阶自相关结构的固定效应部分线性可加动态面板模型中未知参数和函数的估计;进一步,证明了所得估计量的一致性和渐近正态性,同时还用Monte Carlo模拟实验比较了该方法和半参数GMM法在有限样本下的表现;最后将所述方法应用于实际数据分析中.     

基于众数回归的部分函数型线性可加模型的稳健估计

本文讨论部分函数型线性可加模型参数的稳健估计,该模型由经典的可加回归模型和函数型线性模型组合而成.采用B-样条基函数对模型中斜率函数和非参数可加函数进行近似,然后通过最大化众数回归目标函数得到基于众数回归的估计.在一些正则条件下,本文给出估计的收敛速度和渐近分布.最后通过模拟计算和应用实例以表明所提方法的有效性.模拟结果表明,该方法不仅具有稳健性,即不易受污染数据或厚尾分布的影响,而且在信噪比较大时可以与最小二乘方法有相同的表现.     

部分函数型线性可加分位数回归模型

文章结合可加分位数回归模型和函数型线性分位数回归模型,提出了部分函数型线性可加分位数回归模型.我们采用函数型主成分基函数逼近斜率函数,B-样条基函数逼近可加函数,提出了模型的估计方法;在一些基本的假设条件下,给出了斜率函数估计和可加函数估计的收敛速度;最后通过模拟计算和应用实例表明了所提方法的有效性.     

固定效应部分线性单指标面板模型的惩罚二次推断估计

单指标面板模型已广泛应用于各学科领域的研究中,其估计方法较为丰富,然而鲜有估计方法将个体内的相关性考虑在内.基于此,本文研究了一类个体内存在相关性的固定效应部分线性单指标面板模型,采用惩罚二次推断函数法和LSDV法相结合的方法对模型进行估计,证明了所得估计量的一致性和渐近正态性.Monte Carlo模拟结果显示其具有优良的有限样本表现,并将该估计技术应用于实际数据分析中.     

多元部分线性模型的B-样条估计(英文)

本文考虑多元部分线性回归模型的估计问题,得到了该模型参数的最小二乘估计和非参数函数的B-样条估计,并证明了参数估计的渐近正态性,给出了非参数函数估计的最优收敛速度.     

纵向数据下非参数带测量误差的部分线性变系数模型的估计

本文研究纵向数据下非参数部分带有测量误差的部分线性变系数模型的估计.利用B样条函数近似模型中的变系数函数,构造偏差修正的二次推断函数,得到模型中未知参数和变系数函数的估计.证明变系数函数估计量的相合性和参数估计量的渐近正态性.数值模拟和实例分析结果表明所提估计方法在有限样本下的有效性.     

部分线性自回归模型中回归函数的半参数估计（英文）

本文主要研究具有一阶自回归误差的三阶部分线性自回归模型中回归函数的半参数估计问题.假定回归函数来自某个参数分布族,利用条件最小二乘法得到参数估计量,再结合非参数核函数进行调整,给出回归函数的半参数估计量.并在一定条件下,证明了估计量具有相合性.最后,通过模拟研究验证了此方法的有效性.     

带固定效应空间误差面板数据模型的经验似然推断

本文研究了带有固定效应的空间误差面板数据模型的经验似然推断问题.利用经验似然方法,通过鞅差序列将空间误差面板数据模型估计方程中的二次型转化为线性形式,构造了空间误差面板数据模型参数的经验似然比统计量,得到了统计量的极限分布.     

Generalized Partially Linear Measurement Error Models

This article considers generalized partially linear models when the linear covariate is measured with additive error. We propose estimators of parameter and nonparametric function by using local linear regression, the SIMEX technique, and generalized estimating equation. The asymptotic normality of the estimators of the parameter, and bias and variance of the estimators of the nonparametric component are derived under appropriate assumptions. In addition, the generalization to clustered measurements is discussed. The approaches are used to the analysis of data from the Framingham Heart Study. A simulation experiment is conducted for an illustration.     

The FEXP estimator for potentially non-stationary linear time series

We consider semiparametric fractional exponential (FEXP) estimators of the memory parameter d for a potentially non-stationary linear long-memory time series with additive polynomial trend. We use differencing to annihilate the polynomial trend, followed by tapering to handle the potential non-invertibility of the differenced series. We propose a method of pooling the tapered periodogram which leads to more efficient estimators of d than existing pooled, tapered estimators. We establish asymptotic normality of the tapered FEXP estimator in the Gaussian case with or without pooling. We establish asymptotic normality of the estimator in the linear case if pooling is used. Finally, we consider minimax rate-optimality and feasible nearly rate-optimal estimators in the Gaussian case.     

Partially varying coefficient single-index additive hazard models

The partially linear additive hazards model has been proposed to study the interaction between some covariates and an exposure variable. In this paper, we extend it to the partially varying coefficient single-index additive hazard model where the high dimension covariates are collapsed to a single index, due to practical needs. Two sets of estimating equations were proposed to estimate the varying coefficient functions in the linear components: the link function for the single index and the single-index parameter vector separately. It was shown that the proposed local and global estimators are asymptotically normal. Simulation studies were conducted to examine the finite-sample performance of our method to compare the relative performance of our method with existing ones. A real data analysis was used to illustrate the proposed methods.     

Estimation of moments for linear panel data models with potential existence of time effects

In econometric analysis of panel data, one always doesn’t have enough information to assure the existence/absence of time effects, which can lead to wrong conclusions in statistical inference such as moment estimation and hypothesis testing. In this paper, estimation of second and fourth order moments of the individual effects and the errors are studied for linear panel data models without information on the existence/absence of time effects. With differences of the residuals over the individual index, the orthogonality-based moment estimators of the random individual effects and the errors are respectively obtained without affecting each other. These moment estimators are robust on the potential existence of time effects. Their asymptotic normalities are obtained under some moment conditions. Monte Carlo simulations are carried out for illustration.     

Estimations of parametric functions under a system of linear regression equations with correlated errors

Estimations of parametric functions under a system of linear regression equations with correlated errors across equations involve many complicated operations of matrices and their generalized inverses. In the past several years, a useful tool -- the matrix rank method was utilized to simplify various complicated operations of matrices and their generalized inverses. In this paper, we use the matrix rank method to derive a variety of new algebraic and statistical properties for the best linear unbiased estimators （BLUEs） of parametric functions under the system. In particular, we give the necessary and sufficient conditions for some equalities, additive and block decompositions of BLUEs of parametric functions under the system to hold.     

Efficient and fast spline-backfitted kernel smoothing of additive models

A great deal of effort has been devoted to the inference of additive model in the last decade. Among existing procedures, the kernel type are too costly to implement for high dimensions or large sample sizes, while the spline type provide no asymptotic distribution or uniform convergence. We propose a one step backfitting estimator of the component function in an additive regression model, using spline estimators in the first stage followed by kernel/local linear estimators. Under weak conditions, the proposed estimator’s pointwise distribution is asymptotically equivalent to an univariate kernel/local linear estimator, hence the dimension is effectively reduced to one at any point. This dimension reduction holds uniformly over an interval under assumptions of normal errors. Monte Carlo evidence supports the asymptotic results for dimensions ranging from low to very high, and sample sizes ranging from moderate to large. The proposed confidence band is applied to the Boston housing data for linearity diagnosis. Supported in part by NSF awards DMS 0405330, 0706518, BCS 0308420 and SES 0127722.     

An efficient risk estimator with external information under additive hazards model

Rare event data is encountered when the events of interest occur with low frequency, and the estimators based on the cohort data only may be inefficient. However, when external information is available for the estimation, the estimators utilizing external information can be more efficient. In this paper, we propose a method to incorporate external information into the estimation of the baseline hazard function and improve efficiency for estimating the absolute risk under the additive hazards model. The resulting estimators are shown to be uniformly consistent and converge weakly to Gaussian processes. Simulation studies demonstrate that the proposed method is much more efficient. An application to a bone marrow transplant data set is provided.     

利用分层抽样数据中完全辅助信息的模型校正方法

In stratified survey sampling, sometimes we have complete auxiliary information. One of the fundamental questions is how to effectively use the complete auxiliary information at the estimation stage. In this paper, we extend the model-calibration method to obtain estimators of the finite population mean by using complete auxiliary information from stratified sampling survey data. We show that the resulting estimators effectively use auxiliary information at the estimation stage and possess a number of attractive features such as asymptotically design-unbiased irrespective of the working model and approximately model-unbiased under the model. When a linear working-model is used, the resulting estimators reduce to the usual calibration estimator(or GREG).     

纵向数据半参数回归模型的估计方法

本文考虑纵向数据半参数回归模型,通过考虑纵向数据的协方差结构,基于Profile最小二乘法和局部线性拟合的方法建立了模型中参数分量、回归函数和误差方差的估计量,来提高估计的有效性,在适当条件下给出了这些估计量的相合性.并通过模拟研究将该方法与最小二乘局部线性拟合估计方法进行了比较,表明了Profile最小二乘局部线性拟合方法在有限样本情况下具有良好的性质.