抽样误差

An algorithm of error analysis is proposed to avoid the conflict between the precision and the computational burden. The quantitative error of the reliability indices can be given through the error analysis.

论文提出了可靠性评估误差分析的实用算法，为解决蒙特卡洛模拟法精度和抽样次数的矛盾提供了有效的解决途径，为可靠性指标提供了量化误差分析。

In data space, multi-stage anti-aliasing digital filters are designed on the basis of the Nyquist sampling theorem and the minimax criterion to ensure that the observational data are sampled and stored without distortion.

在数存空间内，依据奈奎斯特定理和最大误差最小化的原则，设计多级抗混叠数字滤波器，使实测数据经过滤波器后，达到无失真抽样与存储的目的。

We give the approximation formulas of the variance of the sample ratio estimator for a population mean and its asymptotically non-biased estimator in multi-stage sampling.

讨论多阶段抽样比率估计及其样本量选择的问题，给出了比率估计抽样均方误差的近似公式及其渐近无偏估计的公式。

It proposed a new way to resample time series based on residues.

提出一种新的利用误差项对时序数据样本进行BootStrap重抽样的方法。

Hierarchical linear modeling is the most frequently used technique for analyzing multilevel data. In this paper, the twelve key issues in multilevel research and/or using HLM software were reviewed, and provided their exact meanings and feasible solutions. The 12 key issues included intraclass correlation coefficient, sample sizes, formation of organizational constructs, centering problems, fixed effect and random effect, estimation methods, goodness of fit index, explained variation, multicollinearity, robust standard error, level one variance equation, and empirical Bayes estimates.

本文针对阶层线性模式在多层次研究上有关於技术、测量与方法论所遭遇的问题，进行文献整理并讨论其实务意涵，并依据使用时机、抽样议题、资料汇总、分析方法、与模式评估与估算等五方面整理出十二项重要议题与解决之道，包括：组内相关系数的意义与判断准则、多层次研究各层的样本数问题、变数聚合成组织层次问题、中心化的意义、固定效果与随机效果的设定、估计法的选择、适配度的比较、解释变异量的计算、多元共线性的分析、强韧性标准误、第一层误差项变异数与实证贝氏估计值的运用。

Through adopting this algorithm the timing error is shortened to a half of the sample time interval.

通过采用这种算法可将定时误差缩小到只有抽样时间间隔的一半。

Then the sampling method of network traffic measure and error checking method is introduced.

然后介绍了抽样测量的一般方法，以及抽样测量分析中的误差检验方法。

The traditional methods of testing the degree of compaction only test several points,and these points can hardly represent all the area which had been compacted. So it is essential to develop a speediness,undestructive,quantificational instrument for the measurement of the degree of compaction.

传统的压实质量检测方法是用人工抽样检测来推测所有的压实区域，其结果误差较大，易造成漏检，因此研制开发一种快速、无损、定量的压实度检测仪器迫在眉睫。

The key technologies of Grid Method including grid divided in basin, impact radius of rainfall and number of reference rainfall stations have been discussed with example. The parameters of Grid Method have been selected through trial and error. Using simple random sampling to simulate the failure to collect rainfall, the relation between surface rainfall calculation error and station density, station distribution, uneven spatial distribution of rainfall has been sought.

论文探讨了流域网格剖分、降雨影响半径、搜索半径范围和参证站数量等关键技术，研究了网格法参数的优选，并采用简单随机抽样模拟降雨缺测情况，给出了面雨量计算误差与流域站网密度、均匀度、降雨分布不均匀性指标的关系。

By the time of its fall, most of the prisoners were writers who had written against the corruptions of the government.

到它被攻陷的时候，里面多数的犯人是写了反对政府贪污文章的作家。

The most obvious variation to ovum morphological character was that the color was changed from light green to sepiaceous in embryonic development, and all the ovums were almost hatched after 96h.

在胚胎发育过程中卵的形态特征最明显的变化是颜色从淡绿到深褐色，卵在发育96h后卵基本全部孵化。