离差

10%至15%的离差是可以接受的。

A 10 to 15 percent deviation is considered acceptable .

静态条件下SLK交货期完工时间的离差平方和最小问题

Static problem minimizing the square deviation of job completion times under SLK due-dates

单形分布变离差检验的Score统计量的局部影响

Local Influence on Score Statistic for Varying Dispersion Test in Simplex Distribution

基于连续可分的广义非线性纵向数据模型偏离名义离差的score检验及其功效

Score Tests and Powers for Departures from Nominal Dispersion in Separable and Continuous Generalized Nonlinear Models with Longitudinal Data

而EVA评价方法应尽可能地采取比较客观的方法，如离差最大化法。

EVA performance measure method should be impersonal comparatively .

接着使用离差平方（Ward）聚类分析方法对208个台站进行区域划分，通过对分区结果分析，表明该分区接近真实的太阳辐射分布。

Paper use Ward clustering method to carry out regional partition for 208 stations . By analysing the distribution result is close to actuality .

同时，用主成分分析（PCA）作为排序的方法和系统聚类中的离差平方和法与类平均法，对东灵山植物群落的生活型谱进行了排序和分类。

Ordination and classification of the life form of plant community were done by principle component analysis ( PCA ) and systemic cluster analysis .

该文分别在模型离差（dispersion）的权不变和变异时，讨论了指数族广义非线性随机系数模型的变离差的检验问题，得到了score检验统计量。

This paper deals with this type of models and discusses the tests of varying dispersion for constant and nonconstant weights of dispersion respectively . Several score test statistics are obtained .

Poisson回归模型广泛地应用于分析计数型数据，但该模型往往存在偏大离差（overdispersion）问题。

Poisson regression models are widely used in analyzing count data . However , overdispersion often exists in these models .

刻画Poisson回归模型的偏大离差性的两种方法是拟似然方法和随机效应法（Lee＆Nelder，2000），已有许多作者利用随机效应法研究了Poisson模型的偏大离差的检验问题。

Two ways of modelling overdispersion in Poisson regression models are quasi-likelihood approach and random-effect approach . Many authors have discussed the tests for detecting overdispersion in Poisson models using random-effect methods .

对于多指标综合评估问题，分别采用AHP法、强制评分法、专家咨询法三种主观赋权法以及熵权法、离差分析法两种客观赋权法对指标进行赋权。

Three subjective weighting methods including AHP , mandatory marking method and expert-consulting are applied in multi-index comprehensive assessment problem , as well as two objective weighting methods including entropy weight method and variation coefficient method .

非线性Poisson-Gamma回归模型中存在偏大离差时离差参数的齐性检验

Testing for Homogeneity of Dispersion Parameters under Overdispersion in Nonlinear Poisson-Gamma Regression Models

定性分析表明，ACE方法与以往相似-动力模式原理是等价的，但无需重新建立复杂的相似离差预报模式，更具可行性和业务应用前景。

The qualitative analyse show that , the ACE is theoretically equivalent to the previous analogue-dynamical model principle , but need not rebuild the complicated analogue-deviation model , so has better feasibility and operational foreground .

方法：以专家评议结果的平均绝对离差和平均相对离差为指标，通过K-means聚类方法，对全部评议专家进行分类，对明显偏离整体的专家评议结果赋予权重。

Methods : Setting the mean utter and relative deviation from mean as indexes , the author classified the experts with K-means clustering and endowed weights to the results which deviated the common results .

本文利用人均GDP的标准差和离差来测算青海省46个县域经济的绝对差异和相对差异，分析了1998-2006年青海省区域经济差异的时间变化特征。

This article uses average per person GDP the standard deviation and the deviation surveys the Qinghai Province 46 county territory economy the absolute difference and the relative different , has analyzed 1998-2006 year Qinghai Province regional economies difference time variation characteristic .

根据慢度离差法的基本原理，给出由遗传算法确定AE空间位置、发生时刻及慢度离差5个参量的具体方法。

Based on the fundamental theorem of Slowness Deviation Method , the specific method to determine the source location , occurring time and slowness deviation of an AE by the Genetic Algorithm have been offered in this paper .

最后分别计算49个站点不同历时离差绝对值和平均值、离差平方和平均值以及NSE值平均值。

Then , we can get some mean value , such as sum of absolute value of deviations and sum of squares of deviations and NSE for 49 towns of Fujian ' 9 cities mentioned above .

以人均GDP为指标，采用标准差、离差系数、偏态系数，定量分析了重庆直辖以来区域经济差异的变化，获得了GDP增长率与经济差异之间的关系曲线，并对其进行了分析。

The changes of the regional economic difference of Chongqing are analyzed in the paper with per capita GDP , the standard deviation , coefficient of deviation and coefficient of skewness . Also , the causes and current situation of Chongqing province regional economic are analyzed .

另外经过对组内随机取样法、最大离差度法和最大遗传距离法3种不同抽样方法的比较，认为最大遗传距离法较为适合用RAPD数据构建黄瓜核心样本。

Secondly , three sampling strategies ( random sampling method , maximal deviation method , and maximal genetic distance method ) were compared . The result demonstrated that the strategy of maximal genetic distance was better than others in establishing cucumber core collection with RAPD data .

本文应用二倍体离差指数（2cDI）、五倍体超过率（5cER）和恶性指数（MI）等参数，对肝癌和肝硬变DNA含量进行了分析。

2c deviation index ( 2cDI ) , 5c exceeding rate ( 5cER ) and malignant index ( MI ) were used to analyze the DNA content of tissue in cirrhosis and hepatocellular carcinoma ( HCC ) .

基于这一心理研究结论和极大极小决策准则，建立了证券资产选择的极大极小价值离差模型（MMVD）；

Based on the alove conclusion and minimax decision-making rule , a minimax value-deviations portfolio model ( MMVD ) was put forward .

最后，我们考虑股票市场的实际交易约束（如交易费用、基数约束等），构建具有市场摩擦的机会约束下的均值一半绝对离差投资组合模型，并利用SAA方法对模型进行求解。

In addition , we account for real-world trading constraints , such as transaction costs , cardinality constraint etc. and so the chance-constrained mean semi-absolute deviation portfolio model with market frictions is set up and is solved by sample average approximate method .

本文对方波注入法取得的晶体管电磁脉冲损伤阈值数据进行了分析，采用界限范围标准离差（BSD），进行损伤阈值下限范围估计。

The data of damage threshold of a transistor by electromagnetic pulses were obtained with square wave injection and analysed . An evaluation of the lower limit range of the damage threshold was made using the boundary standard difference ( BSD ) .

首先利用主成分分析法，通过DPS数据处理系统，对15个指标因子进行降维处理，确定出主成分，然后运用系统聚类方法中的离差平方和标准对山东省进行灌溉农业分区。

Make use of the Principal Component Analysis Method to make sure of the principal component , through dealing with 15 indexes by the DPS data processing system , and then make use of the Clustering Analysis to carry on the dividing zone for the agriculture irrigation .

按照不同的市值或行业将股票分为若干投资组合，采用组合预期收益离差（ERD）衡量交易者信念的异化程度，研究其是否包含组合或个股收益波动性的信息。

Firstly , all the stocks in Shanghai Stock Exchange are divided into several portfolios based on capitalizations or industries . Then , it measures heterogeneous beliefs of investors using cross-sectional expected return dispersion ( ERD ) . Lastly , it verifies whether ERD includes the new information about volatility .

笔者用其中4种新策略，即新组内离差平方和法、新组内方差法、新组内平均距离法和加权平均距离法，对德国西北部Hopsten干草地进行了分类。

Four of them , sum of squares of new cluster , variance of new cluster , average distance of new cluster and weighted average distance were applied to the classification of dry grasslands in Hopsten , Germany .

基于离差系数转化法的齿轮传动稳健优化设计

Robust Optimal Design of Gear Driver Based on Coefficient of Variation

离差平方和法在管理信息系统结构设计中的应用

The Application of Derivate Square in Designing the Structure of MIS

基于价值离差的资产选择模型比较分析

Comparison and Analysis of Portfolio Selection Model Based on Value Deviation

基于离差最大化的模糊多属性决策方法

Fuzzy Multiple Attribute Decision Making Method Based on the Maximal Deviation