基于白化值的区间DEA建模

传统数据包络分析要求输入输出数据为精确数,然而在某些实际应用中,区间形式的数据相较于精确数更容易获得.将区间数转化为白化值,并基于传统C~2R模型提出了基于白化值的区间C~2R模型.考虑到决策单元的有效性不易通过基于白化值的区间C~2R模型来判断,因此将非阿基米德无穷小概念引入到上述模型,构建了具有非阿基米德无穷小的区间C~2R模型.此外,还给出了用于判断决策单元有效性的区间目标规划方法:分别通过G_(IC~2R)模型和WG_(IC~2R)模型判断决策单元是否为区间DEA有效与区间弱DEA有效.     

GARCH(1,1)模型的稳健估计比较及应用

首先阐述了GARCH(1,1)模型稳健估计的构造方法,然后在模型有无异常值扩散效应约束和异常值比例不同的情况下,比较了传统QMLE估计和多种稳健M估计的表现,结果表明:在数据无异常值下,QMLE估计较优;随着异常值比例增加,稳健Andrew估计表现更好;模型施加异常值扩散效应约束对估计有一定改善但不显著.最后选取波动程度不同的两个阶段沪深300指数的收益率,用模型拟合进行了实例比较,在波动程度较大时,Andrew估计效果较优,在波动相对平稳时,LAD估计较优.     

Scaling considerations for modelling the in situ vitrification process

Thermal-electric field interactions occur in many areas of engineering. An example is the in situ vitrification process, a relatively new technology for remediating hazardous wastes. Mathematical and physical modelling of this process requires knowledge of the similarity constraints involving thermal- electric fields. This paper addresses this issue and derives the scaling relationships for modelling the remediation process from fundamental principles. Requirements for maintaining temperature and electric potential field similarity between the model and the prototype are determined as well as requirements for maintaining similarity in off-gas generation rates. A scaling rationale for designing reduced scale experiments is presented, and the results are assessed numerically.     

On Local Coefficients for Non-generic Representations of Some Classical Groups

This paper is concerned with representations of split orthogonal and quasi-split unitary groups over a nonarchimedean local field which are not generic, but which support a unique model of a different kind, the generalized Bessel model. The properties of the Bessel models under induction are studied, and an analogue of Rodier's theorem concerning the induction of Whittaker models is proved for Bessel models which are minimal in a suitable sense. The holomorphicity in the induction parameter of the Bessel functional is established. Local coefficients are defined for each irreducible supercuspidal representation which carries a Bessel functional and also for a certain component of each representation parabolically induced from such a supercuspidal. The local coefficients are related to the Plancherel measures, and their zeroes are shown to be among the poles of the standard intertwining operators.     

金融危机下Fama-French多因子模型在中国债券市场的应用

利用近3年的中国债券市场数据,建立了Fama-French多因子模型,并利用多元回归方法进行了比较分析.结果表明,TERM-DEF两因子模型有着较好的适用性,但存在进一步改进的空间.     

一类数列极限的概率求法

通过建立适当的随机模型,应用概率方法可以确定一类具有复杂结构的数列的极限,本文通过举例说明这种方法的运用     

具有椭球等高分布误差的半参数回归模型参数的Bayes估计

本文给出了具有椭球等高分布误差的半参数回归模型中参数的Ｂａｙｅｓ估计．     

基于有限元仿真的东风11型救援小车支撑块结构优化

对东风11型救援小车支撑件进行了有限元分析．通过有限元计算。详细分析了有效应力分布情况；找出了应力集中较大的区域．并分析了其主要原因；观察了应力值最大区域并提出了改善应力集中的方法和措施。从而达到了结构优化的目的．     

汉字识别的加工模型

通过文献分析,概述了认知心理学关于汉字识别过程及其加工机制研究的几种主要的加工模型,即经成分识别模型、多层次交互激活模型、连接主义模型、合体汉字字形识别模型和汉字识别的多层次格式塔双向加工模型.     

SARS传播的数学模型的建立与分析

建立一个SARS传播的实用的、可求解的改进SIR模型,并用差分方法进行求解。利用北京SARS病例数据对模型验证表明,拟合曲线与实际数据基本吻合,平均相对绝对值误差为0．0153。对隔离时机的提前或延后造成的后果进行数值分析,可得出最终确诊病例累计数与隔离时机的延长或提前呈指数增长或下降的结论。