多重线性

On the other hand, partial Least Square Regression which is called the second generation regression method has got good result in handling the multicollinearity in multilinear regression;But under some circumstances the result of such a regression is not satisfactory.

另一方面，偏最小二乘回归方法，被称为第二代回归方法，在处理线性回归中的多重共线问题时，有着较好的效果；但在有些情况下的回归结果却并不理想。

This paper is to mainly discuss the following problems which are diffieult to be solved in commodity freight volume forecast for a period of time: Variable choice and multiple coorelations; Application of the principles of combining quantitative analysis with qualitative analysis; Selection of the remarkable variables by using Multistep Regression Method; Establishment of reasonable freight volume forecast model.

本文针对总货运量预测中较难解决的问题——变量选择和多重共线性，运用定性和定量相结合原则利用逐步回归方法定量挑选显著性变量，建立理想的货运量预测模型。此外，本文在评介货流研究一般方法基础上，根据多地区货物交流目的，探讨了货流预测的方法，提出用Frator法预测货流。

In this paper, we are devoted to establish the multiplicity of positive solutions to superlinear attractive singular equations with Neumann boundary condition...

本论文主要研究具有奇异超线性的Neumann边值问题多重正解存在性问题。

A piecewise linearization data fitting method based on least square principle is used to set the relation function between measuring value and real value,which is then applied to calibrate the channel error in a new multi-calibration method,namely,amplitude calibration,phase calibration and amplitude-phase relation calibration method.

分析了电力线路工频参数测量装置中相差测量对阻抗测量精度的影响，在此基础上提出将基于最小二乘原理的分段线性数据拟合方法应用到装置通道的校准中，实现包括通道幅值、相差及幅值-相差关系的多重校准的新方法。

Based on these and absorption of the foreign company performance results, the thesis combines the theory and the method to analyze the performance of electronic component manufacturing industry enterprises. Based on the analyzing system enterprise performance system in and abroad, the paper raises innovational achievement indexes and evolutional indexes of company performance as to the default of the electronic component manufacturing industry. The paper builds the hierachical clusteing method and bivariate correlations to choose the best evolutional indexes of company performance in order to avoid multicolinearity. The paper proposes the synthetic evaluating method principal component analysis with weighted variable by analytical process. Then, the paper applies the method to analyze the performance of 28 companies of electronic component manufacturing industry.

基于此，在消化吸收国外有关企业绩效评价方面的成果的基础上，针对企业绩效评价的缺陷，本文提出了企业绩效评价的成长性指标和价值性指标；针对改进的AHP与PCA，本文提出了改进的AHP加权主成分分析方法；由于指标之间可能存在着多重共线性，本文运用系统聚类分析与相关分析对企业绩效的成长性指标与价值性指标进行选取；运用改进的AHP加权主成分分析方法与因子分析对电子元器件制造业的28家上市公司进行实证分析，并验证本文提出的综合评价方法的优越性；本文运用灰色模糊评价方法对28家上市公司绩效的非财务指标进行评价；最后，给出相关的结论与建议。

The multiple corresponding analysis is suitable for correlation analysis of big sample categorical variable.

对于中医研究中常见的无序多分类或二分类变量，要同时研究它们之间的联系，用多重对应分析在结果的直观性和可解释性方面优于对数线性模型。

Then based on the analysis of relationship between variances and multi collinear, combined with systematic analysis of literatures and clinical experience, the logisti...

在变量的相关性及多重共线性分析并结合文献的系统分析及临床经验的基础上，应用SAS软件对有显著性意义的危险因素进行回归分析、聚类分析及主成分分析。

Results Collinear variables are excluded according to the above-mentioned result and academic experience.

结果 将去除多重共线性的自变量引入多元回归模型，即可取得比较满意的结果。

In chapter 7, chloride absorption or binding isotherm of concretes was investigated, linear chloride binding capacity and a non-linear coefficient were firstly proposed, based on a large amount of experimental data of concretes under the actions of single, double, or multiple factors in four typical bitterns.

第七章根据不同混凝土在4种典型盐湖的单一、双重和多重因素作用下的大量数据，研究了混凝土对氯离子的吸附/结合规律，提出了线性氯离子结合能力和非线性系数的新概念。

The researchmethodology system is as follows:The variables that can be measured and have significance to stock relative intrinsic value are industry characteristic, profitability,development potential,qulity of earing and asset, the expandable potential of capital stock,the ability to pay off the debt,asset manging ability and the scale of the capital stock;(2)When the MFPM is established, the original variables should be disposed beforehand;(3)As for the research objects.we should value synthesizationly the achievement of the listed,and take the No.

本章所建立的研究方法体系为：（1）上市公司个股基本面因素中可以量化，并且对股票相对内在价值有重要影响的主要有行业特性、盈利能力、成长性、盈利质量及资产质量、股本扩张潜力、偿债能力、资产运营效率和股本规模大小等方面；（2）在建立MFPM模型时，原始指标值是不能直接用于模型计算的，必须对数据进行预处理后才能用于模型中：（3）在确定研究对象时，需要对上市公司经营绩效进行综合评价，在剔除主要靠未来业绩预期定价的第Ⅳ类股后，把第Ⅰ、Ⅱ和Ⅲ殴作为模型研究对象；（4）在选择多元回归模型时，由于原始自变量由于存在错综复杂的关系，尽管对原始自变量进行过预处理，但自变量指标依然会存在一定的多重共线性，为此在建立MFPM模型时，须采用逐步回归模型；（5）在检验MFPM模型有效性时，须从统计意义和投资实战意义两个方面进行。

This one mode pays close attention to network credence foundation of the businessman very much.

这一模式非常关注商人的网络信用基础。

Cell morphology of bacterial ghost of Pasteurella multocida was observed by scanning electron microscopy and inactivation ratio was estimated by CFU analysi.

扫描电镜观察多杀性巴氏杆菌细菌幽灵和菌落形成单位评价遗传灭活率。