残差和

Based on the autoregression equation for the standard variation of the residual errors obtained by ordinary regression analysis, the least square estimates and the maximum likelihood estimates for the regression, the autoregression and the moving average coefficients are derived.

提出一种异方差回归－时序模型，通过建立回归分析残差的标准差自回归方程，给出回归系数、自回归系数和滑动平均系数的最小二乘估计和极大似然估计。

The basic diagnostic statistics such as Cook distance, generalized leverage and residuals et al. are introduced, and the equivalency of case deletion model and mean-shift outlier model is investigated. Some specific problems, such as local influence analysis and the test for heteroscedasticity are discussed. Finally, numerical examples are given to illustrate our diagnostic methods.

首先，证明了均值漂移模型与数据点删除模型的等价性定理并定义了残差、Cook距离等诊断统计量；其次，研究了异常点的Score检验统计量；第三，研究了该模型的局部影响分析，得到了局部影响分析的曲率度量；第四，研究了该模型的异方差检验问题，得到了检验异方差的Score检验统计量和修正的Score检验统计量；最后，通过实例分析验证了这些诊断方法的有效性。

A model for finding the cryptic period of hydrological time series is proposed. The functions of the model include to apply logarithm transformation for making the sequence more stationary, to construct a corresponding auto-regression model for reducing the herteroskedasticity of residual, to determine the component of the prime period by periodogram analysis.

本文在讨论水文时序平稳性和异方差性的基础上，提出了检测水文时序隐含周期的模型，即对水文时序设计对数变换序列，提高序列的平稳性；建立相应的自回归模型降低残差的异方差性；通过对所得平稳序列进行周期图分析确定可能的主周期分量。

Model for finding the cryptic period of hydrological time series is proposed. The functions of the model include to apply logarithm transformation for making the sequence more stationary, to construct a corresponding auto-regression model for reducing the herteroskedasticity of residual, to determine the component of the prime period by periodogram analysis.

bstract 本文在讨论水文时序平稳性和异方差性的基础上，提出了检测水文时序隐含周期的模型，即对水文时序设计对数变换序列，提高序列的平稳性；建立相应的自回归模型降低残差的异方差性；通过对所得平稳序列进行周期图分析确定可能的主周期分量。

Results the parameter estimates and it's stability of linear regression models is associated with the variation of explanatory variable between and within level 2units,and the relationship of parameter estimates and it's stability between linear regression models and variance component model is associated with level 2 residual variance or intatra-unit correlation.

结果 线性回归模型参数估计以及估计的稳定性与自变量在水平2单位间和水平2单位内的变异大小有关，线性因归模型与方差成分模型参数估计及其稳定性的关系与水平2残差方差或单位内相关系数大小有关。

Since in some instance, the variance of response variable might be changeable along withthe alteration of explanatory variable, in the procedure of model constructing, a fixedeffects model should be considered firstly, then the random effects model or randomcoefficient model when necessary. The best fit model should be chosen by comparing thealterations of between-study variance and residual variance, as well as their relationshipswith covariates.

基于各研究中因变量与自变量间的曲线形状可能存在较大差异，建模时可依次考虑固定效应模型、随机效应模型和带协变量的随机系数模型，并观察研究水平的方差与残差方差的变化，以及与协变量的关系，从而选择最优模型。

This thesis,starting from working out the arrest points and the infl ection points of the function curves of the grey mathema tical model GM(1:1 )and its residual correction model and residual period model, derives the conditions for empliying these grey models to obtain a higher class of accuracy,thus enabling the user to decide which model should be used in the light of the state of given initial sequency data.

从求得灰色数学模型GM(1.1)及其残差指数修正模型和残差周期修正模型函数曲线的驻点与拐点入手，导出使用这些灰色模型取得较高精度的条件，让使用者可依据初始序列数据的态势决定采用哪种数学模型，并且在结果未揭示以前便可大体上估计到拟合精度是否令人满意。

Using augmented Dickey-Fuller test, the result shows that BDI logarithm series is nonstationary but the first difference is stationary, i. e. it is integrated of order one. High-level ARCH effect was certification in the BDI logarithm series by ARCH LM test, and GARCH (1, 1) model was used to eliminate the conditional heteroscedasticity. Through variance ratio test, the result shows, that the random walk hypothesis of BDI logarithm series can be rejected and international dry bulk shipping market is inefficient.

运用ADF检验方法对BDI的对数序列进行平稳性和单整检验，结果证明BDI对数序列是一个非平稳过程，经一阶差分后是平稳过程，即BDI对数序列是一阶单整过程；通过ARCH LM检验认为BDI对数序列存在高阶ARCH效应，并用GARCH(1，1)模型消除了残差序列的条件异方差性；通过方差比检验法对国际干散货航运市场的收益率序列进行了检验，结果显示BDI的对数序列的随机游走假设被拒绝，国际干散货航运市场不是一个有效的市场。

On the foundation of standard UKF algorithm, the ratio of the theoretical value to the practical ones of the innovation's covariance was taken as the input of the FLAC to reduce the dependency of FLAC on the filter model, the generality of the fussy adaptive UKF algorithm was intensified; In the procedure of adjusting the measurement noise's covariance dynamically, the adjusted range could be amplified or minified to different extent by setting the exponential adjusting parameter, so the adjusting velocity and precision of the algorithm was controlled.

在标准的非线性UKF算法基础上，以残差的实际方差与理论方差的比值作为FLAC的输入，使FLAC对滤波模型的依赖性减弱，强化了模糊自适应UKF方法的通用性；在对量测噪声方差阵进行动态调节的过程中，通过设置指数调节参数可不同程度地放大或缩小调节的幅度，使算法的调节速度和精度得到控制。

The Auto-RegressiveAR(n Model of time series analysis method is used to study the test.

对原始数列用模型拟合计算得到残差，根据残差的来源和特点，先提取周期性确定项，然后以时间序列分析的AR模型进行研究。

With Death guitarist Schuldiner adopting vocal duties, the band made a major impact on the scene.

随着死亡的吉他手Schuldiner接受主唱的职务，乐队在现实中树立了重要的影响。

But he could still end up breakfasting on Swiss-government issue muesli because all six are accused of nicking around 45 million pounds they should have paid to FIFA.

不过他最后仍有可能沦为瑞士政府&议事餐桌&上的一道早餐，因为这所有六个人都被指控把本应支付给国际足联的大约4500万英镑骗了个精光。