1)  Variance-covariance
方差协方差法
2)  variance-covariance method
方差-协方差法
3)  deviation
方差
1.
After several discrete wavelet transforms, another expression of the deviation of white noise on every scale is given by us.
推导出了经过多次离散小波变换后,白噪声在不同尺度上方差的另外一种表达式,应用到小波域硬阈值滤波中,通过数字仿真发现,二进小波的计算量较大,滤波效果较好,离散小波的计算量较小,滤波效果稍差,并对原因进行了说明。
2.
The author thinks that sampling methods should be chosen according to the characteristics of cases studies and phasing of deducing should be replaced by direct stratificated to optimize the calculation of deviation which can make case deducing studies scientific and more practicable.
本文针对现行限额以下贸易业抽样调查方案在实施过程中反映出来的问题 ,提出了融复合分层和顺序抽选阶段于一体的架构及灵活地选用不同的抽样方法 ,改分阶段推断为直接分层推断 ,优化抽样方差计算等设想 ,目的在于使调查方案更科学、更具有可操作性。
4)  variance
方差
1.
Improving methods to calculate estimator minimum variance of the finite population mean;
总体均值估计量最小方差的改进
2.
Image compression based on the criterion of variance and energy;
基于能量和方差标准的图像压缩方法
3.
A real time fault diagnosis system based on variance of power spectrum at characteristic frequency;
基于特征频率功率谱方差的故障实时检测
5)  square deviation
方差
1.
The mean value and square deviation of residence time measured at different outlet with different stirring speeds and flow rates were calculated and analyzed by L_9(3~3) orthogonal experiments.
用NaCl溶液做示踪剂测定该管式反应器内的停留时间分布(RTD)曲线,并计算了平均停留时间和方差。
2.
Taking the minimum square deviation of the irrigation rate and successive images over the whole irrigation duration as objective function, this model considers 4 kinds of possible adjustment approaches and the corresponding adjustment rules in the actual applications.
该模型以保证灌水率方差在总灌水延续时间内最小和图像连续为目标函数 ,考虑了实际中 4种可能的调整途径及相应的调整规则。
6)  standard deviation
方差
1.
Data expectation,standard deviation,the space between the especially apices,the distance of especially poles and the width of the half especially poles of section data define as factor of this method.
利用地球物理异常显示参数表示剖面数据,以剖面数学期望、方差、特征峰间距、特征峰半极值为信息参数,对放射性测量剖面进行模糊聚类分析。
2.
In order to assess the veracity of scheduled loss ratio,the standard deviation between scheduled loss ratio and practical com-pleting value was utilized to express the precision.
以线损理论计算为基础,并充分考虑站用电、低压配网损耗、管理线损,给出了计划线损率的计算公式,并提出利用计划线损率与实际完成线损率之间的最小方差来衡量其科学性。
参考词条
补充资料:协方差阵


协方差阵
covanance matrix

  协方差阵【cm.dan份ma州x;曰.例...叱幽旧M.,阅a] 若干个随机变量,成对取其协方差,所构成的矩阵.更确切地,k维向量X=(x,,…,习的协方差阵为方阵艺=〔【(火二〔X)(浑‘E幻T],这里〔X=(E戈,…,〔勒丁是均值向量.协方差阵的分量是 aij=日(不一E戈Xxj一Exjll=cov(Xi,xj), i,j=l,…,k,而当i=j时,它与0戈(“var(茂》相同(即戈的方差位犷主对角线_!一).协方差阵是一个对称半正定阵.若协方差阵为正定的则X的分布为非退化的;否则为退化的.对随机向量血言,协方差阵的作用,正如随机变量的方差.如果随机变量X,,…,戈的方差都是1.则X二(刃、,一、戈)的协方差阵与其相关阵(mrrelation matrix)相同. 样本厂”,…,砂、的样本协方差阵,由方差和协方差的估计量构成二 S一汁:户l‘X(用’一见‘X‘”一习了,这里X‘m,如二l,.。)是独立同分布的k维随机向量,而-了是厂,j、…,户’的算术平均.如果丫‘、,二,厂”,的分布是具协方差阵艺的多维且态分布,则S(n一l)/。是艺的最大似然估计量;在这一场合,矩阵(n一飞)S各元的联合分布称为Wi劝斌分布(Wishart distrlbuti(,n).它是多元统计分析中的基本分布之一,借助于它可检验有关协方差阵艺的假设.A.Bfl阳xopoB撰陈希孺译
  
说明:补充资料仅用于学习参考,请勿用于其它任何用途。