广义协方差改进估计的优良性

对半相依回归线性系统,我们周期地使用迭加信息的方法,得出了回归系数广义协方差改进估计。这个估计系列的方差是单调不增的。它改进了回归系数的协方差改进估计。特别当m=2时,我们不对误差向量的分布及设计阵之间的关系作出任何假设,而得到广义协方差改进估计的极限便是BLUE的结果;当协方差阵未知时,得到了两步估计的表达式。     

应用目标规划于自适应回归预测的参数估计

本文应用目标规划和灵敏度分析对自适应回归预测模型的参数进行估计。此种估计参数的方法避免了在最小二乘估计中由于数据“异常”而造成计算结果的偏差。     

非线性计量经济建模变量选择

提出了非线性计量经济建模变量选择的一种方法。采用非参数回归方法确定最佳函数,通过主成分分析方法进行变量筛选,给出了相应的CE算法。仿真实验和计算实例验证了该方法的有效性。     

贝叶斯型的时变系数回归模型

提出了具有时变系数的回归模型及其模型识别和参数估计的贝叶斯方 法．在此模型中对于时变回归系数的变动，应用了高斯型概率差分方程式作 为约束条件，称之为高斯型平滑性事先分布．模型中的超参数（ｈｙｐｅｒｐａｒａｍｅ－ ｔｅｒ）的估计，采用了最大似然估计法．模型的识别（差分次数的决定）应用了 Ａｋａｉｋｅ的最小 ＡＢＩＣ法．给出了模型估计的算法及其应用例子．最后，讨论了 平滑性事先分布中参数的最优估计的意义．     

三波长一元线性回归分光光度法的研究与应用

本文提出了三波长一元线性回归分光光度法。该法将几组不同组成的已知混合溶液,在三个测试波长下所测得的吸光度直接进行线性拟合,并建立了三波长线性方程 y=Kx+B。这种方法较好地克眼了三波长法过于依赖仪器波长精度和准确性的缺陷,以及通常三波长法测定实际混合样品时,完全由纯组分所建立的定量公式而引入的误差。提高了测定精度和可靠性,操作简便易行实用,又能兼顾三波长法的灵敏度和原有特长。     

AN INVESTIGATION OF THE RELATION BETWEEN COTTON FIBRE PROPERTIES AND YARN QUALITY BY USING HIGH VOLUME INSTRUMENT(HVI)SYSTEM

The objective of this paper is to establish reliable prediction equations relating cotton fibreproperties measured by HVI system and yarn quality,A useful statistical method is adopted for de-veloping a multiple regression model interpreting the relation between the data of Spinlab HVIfibre properties and quality parameters of yarn STQ (Strength Tex Quotient).The percent relativecontribution of a fibre property with respect of STQ is also assessed.The results show that the totalcontribution of the HVI measured fibre properties can account for 77.4% of known variation ofyarn STQ.The main feature of the approach is its flexibility in accommodating all fibre properties.The examination of regression equation showed that it could be well applied to predict STQ ofyarns spun from the same spinning system,but,for different spinning systems and also for yarns ofdifferent linear density,modification of the equation would be needed.     

线性模型信息准则的稳健性

在不需要假定：１）被选模型包含真实模型；２）误差分布为正态分布的条件下，应用作者定义的广义Ｋ－Ｌ差异度，得到了自变量选择的广义Ｋ－Ｌ差异度准则。这个准则包含了所有信息准则作为它的特例，由此说明信息准则具有稳健性。     

Time series forecasting using robust regression

The method of ordinary least squares (OLS) and generalizations of it have been the mainstay of most forecasting methodologies for many years. It is well-known, however, that outliers or unusual values can have a large influence on least-squares estimators. Users of automatic forecasting packages, in particular, need to be aware of the influence that outlying data values can have on statistical analyses and forecasting results. Robust methods are available to modify least-squares procedures so that outliers have much less influence on the final estimates; yet these formal methods have not found their way into general forecasting procedures. This paper provides a case study in which classical least-square-estimation procedures are complemented with a robust alternative to enhance statistical fit criteria and improve forecasting performance. The study suggests that much can be gained in understanding the nature of outliers and their influence on forecasting performance by performing a robust regression in addition to OLS.