多元线性模型回归系数估计在矩阵损失下的可容许性

推广了吴启光关于一元线性模型回归系数估计的可容许性的结果，得到了多元线性模型回归系数的估计可容许的充要条件。     

对"佯谬回归系数"问题的探讨

对佯谬回归系数的概念进行了统计意义上的界定,指出了产生佯谬回归系数的6种主要原因,在此基础上,对如何正确理解含有佯谬回归系数的回归预测模型问题提出了以下的观点：应根据产生佯谬回归系数的原因,区别看待含佯谬回归系数的预测模型;含佯谬回归系数的回归模型未必都不科学,在某些特定条件下,有些含佯廖系数的回归预测模型可科学合理地反映研究对象的客观变化规律。     

线性回归模型参数LS估计相合性的一个结果

本文研究了一元线性回归模型中常数项与回归系数的LS估计相合性之间关系，得到一个相合性结果。     

影响废水排放因素的统计分析

蔡越江．影响废水排放因素的统计分析．本文通过定量分析人口数量、能源消费、生活水平、能源水平和废水治理水平等因素对废水排放的直接和间接影响,以及这些因素之间的相互影响和相互作用,结果表明人口数量和工业生产是影响废水排放的重要因素,提高生产率和能源利用率能有效减少废水的排放     

线性回归系数的修约规则

线性回归系数的修约规则黄映辉（西安交通大学管理学院）线性回归根据存有相互关系的变量ｙ，ｘｔ的观测值来确定回归系数ａ和ｂｉ，从而建立线性回归方程。因为观测值ｙ和ｘｉ通常为近似数，所以由ｙ和ｘｉ计算而得的回归系数ａ和ｂｉ也必为近似数，其修约误差可由ｙ，ｘ...     

回归系数Pitman估计的优良性(Ⅱ)

对线性回归模型Ｙ＝Ｘβ＋ε，证明了回归系数β的Ｐｉｔｍａｎ估计的渐近有效性，并推广了Ｐｏｒｔ和Ｓｔｏｎｅ［５］关于位移参数的有关结果，去掉了及文［４］中所施加的矩限制．     

EMPIRICAL LIKELIHOOD FOR LINEAR MODELS UNDER m-DEPENDENT ERRORS

In this paper，the empirical likelihood confidence regions for the regression coefficient in a linear model are constructed under m-dependent errors. It is shown that the blockwise empirical likelihood is a good way to deal with dependent samples.     

灰铸铁石墨形态的神经网络识别研究

在分析分形维、纹理粗细度及二维回归系数共同组成的纹理特征基础上,本文运用人工神经网络方法实现了对灰铸铁石墨形态的识别．相对于传统方法,该方法识别速度快、重现性强、数据更为客观准确,避免了由于主观判断所引起的人为误差．在识别过程中,本文使用了一种确定隐层节点数的新方法,结果显示良好,对于人工神经网络中隐层节点数确定难的问题具有很好的实用参考价值．     

Additive hazards regression under generalized case-cohort sampling

Case-cohort design usually requires the disease rate to be low in large cohort study,although it has been extensively used in practice.However,the disease with high rate is frequently observed in many clinical studies.Under such circumstances,it is desirable to consider a generalized case-cohort design,where only a fraction of cases are sampled.In this article,we propose the inference procedure for the additive hazards regression under the generalized case-cohort sampling.Asymptotic properties of the proposed estimators for the regression coefcients are established.To demonstrate the efectiveness of the generalized case-cohort sampling,we compare it with simple random sampling in terms of asymptotic relative efciency.Furthermore,we derive the optimal allocation of the subsamples for the proposed design.The fnite sample performance of the proposed method is evaluated through simulation studies.     

Composite hierachical linear quantile regression

Multilevel （hierarchical） modeling is a generalization of linear and generalized linear modeling in which regression coefficients are modeled through a model, whose parameters are also estimated from data. Multilevel model fails to fit well typically by the use of the EM algorithm once one of level error variance （like Cauchy distribution） tends to infinity. This paper proposes a composite multilevel to combine the nested structure of multilevel data and the robustness of the composite quantile regression, which greatly improves the efficiency and precision of the estimation. The new approach, which is based on the Gauss-Seidel iteration and takes a full advantage of the composite quantile regression and multilevel models, still works well when the error variance tends to infinity, We show that even the error distribution is normal, the MSE of the estimation of composite multilevel quantile regression models nearly equals to mean regression. When the error distribution is not normal, our method still enjoys great advantages in terms of estimation efficiency.