累积Logistic回归模型结构变点的序贯检验

分类时间序列在生物医学、社会学和遗传学等领域有着广泛的应用,累积Logistic回归模型是分类时间序列建模的一类重要模型.本文基于偏似然得分过程(Partial likelihood score process)提出一种变点序贯检验方法,监测累积Logistic回归模型的结构是否发生变化.原假设下推导检验统计量的极限分...     

GARCH模型变点的Ratio检验

研究GARCH模型参数变点的Ratio检验.首先构造了基于残量累积平方和的Ratio统计量,推导了原假设下统计量的极限分布,其次采用Monte Carlo方法检验其有效性,最后以数据为例进一步说明该方法的实用性.     

随机排列方法检验方差变点

给出了随机排列的主要性质及证明。构造了随机排列检验方差变点的统计量。以GARCH(1,1)过程为例,模拟比较了随机排列方法与近似极限分布方法关于方差变点检验的临界值。应用随机排列方法检测人民币兑美元汇率的变点,并与惩罚对比函数方法作比较。模拟与实证结果均表明随机排列方法检验方差变点是灵活有效的。     

测量误差模型只有一个变点的检验和估计

本文讨论了测量误差模型中参数只有一个变点的检验和估计问题，首先，给出其似然比检验统计量，然后，基于最小信息准则的原理，利用Schwarz信息准则（SIC），在多余参数已知和未知的情况下，分别给出了检验统计量，讨论了利用SIC方法给出的检验统计量的渐近分布，证明了基于似然比方法和SIC方法给出的变点估计是相同的，并且在一定条件下，给出了变点估计的极限分布，运用Monte-Carlo随机模拟的方法，分别给出了以上检验的临界值。     

至多一个变点模型的统计推断

对至多一个变点模型,其中‘(1/n),…,‘(n/n)独立同分布,借助高斯过程理论,利用第一型极值分布逼近本文所提出的变点估计量的分布,讨论了关于变点t_0,跳跃度(α2-α1)和斜率变度(β2-β1)的假设检验和区间估计问题。     

随机设计下非参数回归模型方差变点Ratio检验

研究随机设计下非参数回归模型方差变点Ratio检验.首先用局部多项式方法估计回归曲线得到残差序列,其次基于残差的平方序列构造Ratio检验统计量并推导检验统计量的极限分布.最后数值模拟与实例分析结果表明方法的有效性.     

基于变点模型的我国煤炭价格波动特征研究

分析了我国煤炭价格形成机制的长期演变趋势,利用秦皇岛大同优混煤2003年3月至2015年3月每周的平仓价格数据,采用结构变点检验分析法,实证研究了煤炭价格的波动规律.研究结果表明:煤炭价格呈现出显著的强持久性、高波动性和高强度跳跃性的市场特征.当存在突发事件时,变点模型能较好地拟合煤炭价格的变动过程.注意到,煤炭价格波动性主要由市场供需关系所决定,预计煤炭供大于求的阶段性状态仍会延续,煤炭价格将持续走低.最后提出了相应的对策与建议.     

位置参数模型变点问题的非参数检验

本文引入Fisher-Yates提出的计分函数和Van der Waerden提出的计分函数,对只有一个变点的位置参数模型的假设检验问题分别给出了四个检验统计量.利用Monte-Carlo随机模拟的方法,求出了检验的渐近临界值,并且对本文提出的检验,以及Pettitt在文[1] 中提出的检验,Schechtman和Wolfe在文[2]中提出的检验的势进行了比较。     

测量误差模型参数变点的非参数检验

本讨论测量误差参数变点的检测问题,利用秩统计量,给出了模型只有一个变点的检验统计量,运用检验统计量渐近分布的性质,给出了一个计算检验淅近临界值的公式,由此我们可以较为客易计算检验的临界值。     

ARCH模型分布变点监测与应用

分布变点监测是时间序列交点分析的一个重要内容.为将分布交点监测从线性时间序列模型拓展到非线性时间序列模型,提出一种经验特征函数型的统计量监测ARCH模型误差项平方的分布变点,给出了监测统计量在原假设下的极限分布,并证明了此方法的一致性,用Bootstrap重抽样方法获得了极限分布的临界值,并和Kolmogorov-smirnov型监测统计量进行了比较.模拟结果和实例分析说明了当已观测样本量较大时,采用经验特征函数型统计量监测效果较好.     

时间序列中方差的结构变点的小波识别(英文)

本文给出了时间序列中方差的小波系数的两种估计:连续估计和离散估计.这两种估计可以用来检测时间序列中方差的结构变点.利用这两种估计我们给出了方差变点的位置和跳跃幅度的估计,并且显示出这些估计可达到最佳收敛速度.同时,我们还给出了这些估计的收敛速度以及检验统计量的渐进分布!     

Fixed Points in Tense Models

We study into definability of least fixed points in tense logic. It is proved that least fixed points of tense positive -operators are definable in transitive linear models. Examples are furnished showing that the least fixed points of tense positive operators may fail to be definable in the class of finite linearly ordered models, and the class of finite strictly linearly ordered models. Moreover, in dealing with the modal case, we point out examples of the non-definable inflationary points in the model classes mentioned.     

Detecting Change Points in Polynomial Regression Models with an Application to Cable Data Sets

In this paper, the Schwarz Information Criterion (SIC) is used to detect the change points in polynomial regression models. Switching quadratic regression models with same amount of model deviation and switching polynomial regression models with different amount of model deviation for different segments of regression are considered. The number of separate regimes and their corresponding regression orders are assume to be known. The method is then applied to cable data sets and the change points are successfully detected.     

Dynamic Detection of Change Points in Long Time Series

We consider the problem of detecting change points (structural changes) in long sequences of data, whether in a sequential fashion or not, and without assuming prior knowledge of the number of these change points. We reformulate this problem as the Bayesian filtering and smoothing of a non standard state space model. Towards this goal, we build a hybrid algorithm that relies on particle filtering and Markov chain Monte Carlo ideas. The approach is illustrated by a GARCH change point model.     

Estimation of Parameters of Partially Linear Errors-in-variables Models with Replicated Net Points of Observation

A kind of partially linear errors-in-variables models with replicated net points of observation are studied in this paper. Estimators of unknown parameters are given. Under certain regular conditions, it is shown that the estimators of the unknown parameters are strongly consistent and their a.s. convergence rates are achieved.     

On Modeling Change Points in Non-Homogeneous Poisson Processes

Failures in repairable systems are often described by means of non-homogeneous Poisson processes, identified by their intensity and mean value functions. Intervention on the systems are likely to modify their reliability, and changes in intensities and mean value functions are therefore induced. We consider different scenarios in which interventions take places and propose models describing each of them. Bayesian analyses, relying on Markov-chain Monte Carlo methods, are illustrated along with applications to simulated and real, widely-known, data.     

Structural Change Models with an Application in Cryogenic Thermometry

In a linear regression framework, structural change models are proposed for the detection of abrupt changes in parameter values. Two models are discussed: 1) the pure structural change model, where all the components of the parameter are allowed to change and are tested all together, and 2) the partial structural change model, where only some of the parameter β components might change. For an on-line implementation, a sliding window algorithm is introduced. The procedure was successfully applied to phase transition identification in cryogenic thermometry. *Work partially funded under EU SofTools_MetroNet Contact N. G6RT-CT-2001-05061.     

Saddle Points and Pareto Points in Multiple Objective Programming

In this paper relationships between Pareto points and saddle points are studied in convex and nonconvex multiple objective programming. The analysis is based on partitioning the index sets of objectives and constraints and splitting the original problem into subproblems having a special structure. The results are based on scalarizations of multiple objective programs and related linear and augmented Lagrangian functions. In the nonconvex case, a saddle point characterization of Pareto points is possible under assumptions that guarantee existence of Pareto points and stability conditions of single objective problems. Essentially, these conditions are not stronger than those in analogous results for single objective programming.This research was partially supported by ONR Grant N00014-97-1-784AMS Subject Classification: 90C29, 90C26     

多目标规划的ak—较多有效点与ak—较多最优点

在[1]中,作引入了多目标规划的较多有效点及较多最优点的概念,并讨论了它们的性质,本首次提出了ak-较多有效点与ak-较多最优点的概念,并讨论了ak-较多有效点,ak-较多最优点、ak-较多有效解,ak-较多最优解的相关性质。