面板数据分位回归中适应性LASSO调节参数的选择

在带有罚函数的变量选择中,调节参数的选择是一个关键性问题,但遗憾的是,在大多数文献中,调节参数选择的方法较为模糊,多凭经验,缺乏系统的理论方法.本文基于含随机效应的面板数据模型,提出分位回归中适应性LASSO调节参数的选择标准惩罚交叉验证准则(PCV),并讨论比较了该准则与其他选择调节参数的准则的效果.通过对不同分位点进行模拟,我们发现当残差E来自尖峰分布和厚尾分布时,该准则能更好地估计模型参数,尤其对于高分位点和低分位点而言.选取其他分位点时,PCV的效果虽稍逊色于Schwarz信息准则,但明显优于A1kaike 信息准则和交叉验证准则.且在选择变量的准确性方面,该准则比Schwarz信息准则、Akaike信息准则等更加有效.文章最后对我国各地区多个宏观经济指标的面板数据进行建模分析,展示了惩罚交叉验证准则的性能,得到了在不同分位点处宏观经济指标之间的回归关系.     

发散维数SICA惩罚Cox回归模型的一种修正BIC调节参数选择器（英文）

本文研究了发散维数SICA惩罚Cox回归模型的调节参数选择问题,提出了一种修正的BIC调节参数选择器.在一定的正则条件下,证明了方法的模型选择相合性.数值结果表明提出的方法表现要优于GCV准则.     

对部分线性模型用惩罚最小二乘估计时最优光滑化的注记

部分线性模型也就是响应变量关于一个或者多个协变量是线性的, 但对于其他的协变量是非线性的关系\bd 对于部分线性模型中的参数和非参数部分的估计方法, 惩罚最小二乘估计是重要的估计方法之一\bd 对于这种估计方法, 广义交叉验证法提供了一种确定光滑参数的方法\bd 但是, 在部分线性模型中, 用广义交叉验证法确定光滑参数的最优性还没有被证明\bd 本文证明了利用惩罚最小二乘估计对于部分线性模型估计时, 用广义交叉验证法选择光滑参数的最优性\bd 通过模拟验证了本文中所提出的用广义交叉验证法选择光滑参数具有很好的效果, 同时, 本文在模拟部分比较了广义交叉验证和最小二乘交叉验证的优劣.     

生长曲线模型的惩罚最小二乘估计

主要考虑了生长曲线模型中的参数矩阵的估计.首先基于Potthoff-Roy变换后的生长曲线模型,采用不同的惩罚函数:Hard Thresholding函数,LASSO,ENET,改进LASSO,SACD给出了参数矩阵的惩罚最小二乘估计.接着对不做变换的生长曲线模型,直接定义其惩罚最小二乘估计,基于Nelder-Mead法给出了估计的数值解算法.最后对提出的参数估计方法进行了数据模拟.结果表明自适应LASSO在估计方面效果比较好.     

一类零膨胀半参数混合效应模型的惩罚似然估计

本文将半参数线性混合效应模型推广应用到一类具有零膨胀的纵向数据或集群数据的研究中,提出了一类新的半参数混合效应模型,然后利用广义交叉核实法选取光滑参数,通过最大惩罚似然函数方法与EM算法给出了模型参数部分与非参数部分的估计方法,最后,通过模拟和实例说明了本文方法的有效性.     

带有不可忽略缺失数据的半参数再生散度模型的贝叶斯分析

半参数再生散度模型是再生散度模型和半参数回归模型的推广,包括了半参数广义线性模型和广义部分线性模型等特殊类型.讨论的是该模型在响应变量和协变量均存在非随机缺失数据情形下参数的Bayes估计和基于Bayes因子的模型选择问题.在分析中,采用了惩罚样条来估计模型中的非参数成分,并建立了Bayes层次模型;为了解决Gibbs抽样过程中因参数高度相关带来的混合性差以及因维数增加导致出现不稳定性的问题,引入了潜变量做为添加数据并应用了压缩Gibbs抽样方法,改进了收敛性;同时,为了避免计算多重积分,利用了M-H算法估计边缘密度函数后计算Bayes因子,为模型的选择比较提供了一种准则.最后,通过模拟和实例验证了所给方法的有效性.     

半参数线性混合效应模型的联合变量选择

多数基于线性混合效应模型的变量选择方法分阶段对固定效应和随机效应进行选择,方法繁琐、易产生模型偏差,且大部分非参数和半参数的线性混合效应模型只涉及非参数部分的光滑度或者固定效应的选择,并未涉及非参变量或随机效应的选择。本文用B样条函数逼近非参数函数部分,从而把半参数线性混合效应模型转化为带逼近误差的线性混合效应模型。对随机效应的协方差矩阵采用改进的乔里斯基分解并重新参数化线性混合效应模型,接着对该模型的极大似然函数施加集群ALASSO惩罚和ALASSO惩罚两类惩罚,该法能实现非参数变量、固定效应和随机效应的联合变量选择,基于该法得出的估计量也满足相合性、稀疏性和Oracle性质。文章最后做了个数值模拟,模拟结果表明,本文提出的估计方法在变量选择的准确性、参数估计的精度两个方面均表现较好。     

带有不可忽略缺失数据的广义部分线性模型的贝叶斯分析

广义部分线性模型是广义线性模型和部分线性模型的推广,是一种应用广泛的半参数模型.本文讨论的是该模型在线性协变量和响应变量均存在非随机缺失数据情形下参数的Bayes估计和基于Bayes因子的模型选择问题,在分析过程中,采用了惩罚样条来估计模型中的非参数成分,并建立了Bayes层次模型;为了解决Gibbs抽样过程中因参数高度相关带来的混合性差以及因维数增加导致出现不稳定性的问题,引入了潜变量做为添加数据并应用了压缩Gibbs抽样方法,改进了收敛性;同时,为了避免计算多重积分,利用了M-H算法估计边缘密度函数后计算Bayes因子,为模型的选择比较提供了一种准则.最后,通过模拟和实例验证了所给方法的有效性.     

连续型半参数广义线性纵向数据模型的方差成分检验

回归模型的方差成分检验是一个非常重要的问题.该文针对离差参数的变异, 随机效应的影响及两者同时具有的三种情形, 研究了基于纵向数据的连续型半参数广义线性模型的方差成分检验, 得到了Score检验统计量, 最后通过计算机模拟验证了该文所提出的方法的有效性.     

高维模型选择方法综述

模型选择是统计学的热点研究问题。近年来随着数据维数越来越高,传统模型选择方法的应用受到了很多制约。本文着重介绍高维模型选择的新方法,并讨论实现模型选择过程的一个重要环节,即调整参数的选取。最后文章总结归纳了未来可能的研究方向。     

Feature extraction using two-dimensional local graph embedding based on maximum margin criterion

In this paper, we propose a novel method for image feature extraction, namely the two-dimensional local graph embedding, which is based on maximum margin criterion and thus not necessary to convert the image matrix into high-dimensional image vector and directly avoid computing the inverse matrix in the discriminant criterion. This method directly learns the optimal projective vectors from 2D image matrices by simultaneously considering local graph embedding and maximum margin criterion. The proposed method avoids huge feature matrix problem in Eigenfaces, Fisherfaces, Laplacianfaces, maximum margin criterion (MMC) and inverse matrix in 2D Fisherfaces, 2D Laplacianfaces and 2D Local Graph Embedding Discriminant Analysis (2DLGEDA) so that computational time would be saved for feature extraction. Experimental results on the Yale and the USPS databases show the effectiveness of the proposed method under various experimental conditions.     

An efficient method for feature selection in linear regression based on an extended Akaike’s information criterion

A method for feature selection in linear regression based on an extension of Akaike’s information criterion is proposed. The use of classical Akaike’s information criterion (AIC) for feature selection assumes the exhaustive search through all the subsets of features, which has unreasonably high computational and time cost. A new information criterion is proposed that is a continuous extension of AIC. As a result, the feature selection problem is reduced to a smooth optimization problem. An efficient procedure for solving this problem is derived. Experiments show that the proposed method enables one to efficiently select features in linear regression. In the experiments, the proposed procedure is compared with the relevance vector machine, which is a feature selection method based on Bayesian approach. It is shown that both procedures yield similar results. The main distinction of the proposed method is that certain regularization coefficients are identical zeros. This makes it possible to avoid the underfitting effect, which is a characteristic feature of the relevance vector machine. A special case (the so-called nondiagonal regularization) is considered in which both methods are identical.     

自适应稀疏伪谱逼近新方法

自适应稀疏伪谱逼近法是广义混沌多项式类方法的最新进展,相对于其它方法具有计算精度高、速度快的优点.但它仍存在如下缺点：1）终止判据对逼近误差的估计精度偏低;2）只适用于单输出问题.本文提出了适用于多输出问题且具有更高逼近精度的自适应稀疏伪谱逼近新方法.本文首先提出了新型终止判据及基于此新型终止判据的自适应稀疏伪谱逼近新方法,并以命题的形式证明了新型终止判据相比于现有终止判据具有更高的估计精度,从而使基于此的逼近函数精度更接近于预期精度;进而,本文基于指标集的统一策略和新型终止判据,提出了适用于多输出问题的自适应稀疏伪谱逼近新方法,该方法因能充分利用各输出变量的抽样结果,具有比将单输出方法直接推广到多输出问题更高的计算效率.多个算例验证了本文所提出新方法的有效性和正确性.     

基于概率语言BWM与PROMETHEEII的多准则决策方法

提出了一种改进的偏好顺序结构评估法(Preference Ranking Organization Method For Enrichment Evaluations, PROMETHEE),即基于概率语言BWM与PROMETHEE II的多准则决策方法。针对多准则决策过程中,评价信息的模糊性和不确定性,采用概率语言(Probabilistic Linguistic Term Set, PLTS)处理评价信息,并纳入PROMETHEE II中对备选方案进行排序。针对传统的PROMETHEE II 中准则权重需要从外部获得的问题,采用最优最劣法(Best-worst Method, BWM)确定准则的权重。最后,以某企业无人机改造方案为例进行分析,验证所提方法的有效性和可行性。     

Generalized criterion of the long-term strength of viscoelastic materials

A long-term strength condition of the same type as the Il'yushin [2] or Moskvitin [1] nonlinear criterion is proposed. A method is suggested for determining the stress and time functions forming part of the proposed criterion from the long-term strength diagrams and from the data of pulsed or stepped loading tests. It is shown, on the basis of the experimental data, that the proposed criterion, which in a particular case is identical with the Moskovitin criterion, describes the experimental results better than the Bailey criterion.     

Sequential change-point detection for mixing random sequences under composite hypotheses

The problem of sequential detection of a change-point in the density function of one-dimensional distribution of observations from a mixing random sequence is considered when both before and after a change-point this density function belongs to a certain family of distributions, i.e. in the situation of composite hypotheses. A new quality criterion for change-point detection is proposed. The asymptotic a priori lower bound for this criterion is proved for wide class of methods of change-point detection. An asymptotically optimal method of change-point detection is proposed for which this lower bound is attained asymptotically. In particular, for the case of a simple hypothesis before a change-point, this method coincides with the generalized cumulative sums (CUSUM) method.      

LMI Optimization Approach to Synchronization of Stochastic Delayed Discrete-Time Complex Networks

Complex networks are widespread in real-world systems of engineering, physics, biology, and sociology. This paper is concerned with the problem of synchronization for stochastic discrete-time drive-response networks. A dynamic feedback controller has been proposed to achieve the goal of the paper. Then, based on the Lyapunov second method and LMI (linear matrix inequality) optimization approach, a delay-independent stability criterion is established that guarantees the asymptotical mean-square synchronization of two identical delayed networks with stochastic disturbances. The criterion is expressed in terms of LMIs, which can be easily solved by various convex optimization algorithms. Finally, two numerical examples are given to illustrate the proposed method.     

Regional eigenvalue-clustering robustness of linear output feedback systems with both time-varying structured and unstructured uncertainties as well as nonlinear actuators

This paper investigates the regional eigenvalue-clusteringrobustness of linear output feedback systems with both time-varyingstructured (elemental) and unstructured (norm-bounded) parameteruncertainties as well as nonlinear actuators. A sufficient conditionis proposed for ensuring that all the eigenvalues of outputfeedback systems with both time-varying structured and unstructuredparameter uncertainties as well as nonlinear actuators remaininside the specified region. No restriction is imposed on theshapes of the specified region. The proposed method is applicableto both the continuous-time case and the discrete-time case.When all the eigenvalues are just required to locate in thestable region, the proposed criterion will become the stabilityrobustness criterion. Two numerical examples are given to illustratethe application of the proposed sufficient condition.     

Consensus protocol design for discrete‐time networks of multiagent with time‐varying delay via logarithmic quantizer

This article considers the problem of consensus for discrete‐time networks of multiagent with time‐varying delays and quantization. It is assumed that the logarithmic quantizer is utilized between the information flow through the sensor of each agent, and its quantization error is included in the proposed method. By constructing a suitable Lyapunov‐Krasovskii functional and utilizing matrix theory, a new consensus criterion for the concerned systems is established in terms of linear matrix inequalities (LMIs) which can be easily solved by various effective optimization algorithms. Based on the consensus criterion, a designing method of consensus protocol is introduced. One numerical example is given to illustrate the effectiveness of the proposed method. © 2014 Wiley Periodicals, Inc. Complexity 21: 163–176, 2015     

Improved stability criterion and output feedback control for discrete time-delay systems

This paper investigates the stability and stabilization for a class of linear systems with time-varying delay. We provide a new finite-sum inequality which is a powerful tool for stability analysis of time-delay systems. Applying the inequality, a new stability criterion is proposed in terms of linear matrix inequalities (LMIs). We also design a method for static output feedback (SOF) control problems which contains two parts. The first part is to find an initial values of the matrix variables. By utilizing the initial values, the condition for SOF control problems can be solved by an improved path-following method. Numerical examples demonstrate the effectiveness of the stability criterion and the SOF stabilization method.