NA样本半参数回归模型估计的矩相合性

考虑了误差为NA序列的半参数回归模型,利用非参数估计方法给出了模型参数的最小二乘估计和加权最小二乘估计,并在适当条件下得到了它们的矩相合性.     

不等式约束条件下部分线性回归模型的参数估计

本文研究了不等式约束条件下部分线性回归模型的参数估计问题，利用最优化方法和贝叶斯方法，给出了不等式约束条件下部分线性回归模型的最小二乘核估计和最佳贝叶斯估计，并且证明了在一定条件下，带约束条件的最小二乘核估计在均方误差意义下要优于无约束条件的最小二乘核估计。     

半参数回归模型的一类压缩估计

研究了半参数回归模型的参数估计问题,利用压缩估计方法给出了模型的一类有偏估计,并与最小二乘估计、岭估计、几乎无偏岭估计进行了比较.在均方误差意义下,新的压缩估计明显优于最小二乘估计.最后讨论了有偏参数选取的问题.     

误差分布未知下时空模型的自适应非参数估计

极大似然估计作为参数估计中较为有效的一种估计方法,在误差分布未知下无法进行, 另一方面, 时空数据经常含有奇异点或来自重尾分布,此时基于最小二乘的估计方法效果欠佳.考虑时空异质性和相关性,针对误差分布未知的时空模型,本文提出基于核密度估计的自适应非参数估计方法.在较弱的条件下证明了该估计量和已知误差分布下的局部极大似然估计量是渐近等效,比基于最小二乘的局部多项式估计量有效. 模拟和实证都验证了该方法对于有限样本的有效性, 尤其奇异点的存在,该方法在边界的拟合效果显著优于基于最小二乘的方法.     

相依误差下非线性回归模型LS估计的收敛速度

研究非线性模型的参数估计问题,在误差满足较宽泛的条件时,证明了参数的最小二乘估计具有强相合性及强相合速度.     

确定Lotka-Volterra生态系统模型高精度参数的研究

研究确定Lotka-Volterra生态系统模型的高精度参数估计问题.利用周期性,先对测量数据进行预处理;然后用三种不同的方法构造了误差函数,进行非线性最小二乘法参数估计;再用计算机仿真对其进行验证.结果表明该方法能够有效地解决高精度参数估计中消除测量数据误差的问题.     

系数为梯形模糊数的模糊回归分析的最小二乘法

由于模糊数往往可以用梯形模糊数来逼近,因此对梯形模糊数的模糊回归模型的研究就有一定的实用价值.采用最小二乘的方法,针对输入为精确数、输出和回归系数都是梯形模糊数的模糊线性回归模型,讨论了该模型回归系数的最小二乘估计及误差项的估计,实例说明了提出的参数估计的拟合度比较好.     

变系数EV模型基于核估计的误差方差估计

在利用核函数法和广义最小二乘法讨论变系数EV模型系数参数估计的基础上给出了其误差方差σ2的一种估计量(σ)2n,证明了所定义的估计量(σ)2n有很好的大样本性质.     

线性模型中广义最小二乘参数估计误差分布的稳健性研究

研究了线性模型中广义最小二乘参数估计的误差分布稳健性问题.首先讨论了在线性统计模型里,设计矩阵为列降秩矩阵时,模型中给出了误差最大分布类,当误差向量的分布在此范围内变动时,LS估计和GLS估计在均方误差矩阵准则下是最优估计.然后进一步探讨广义最小二乘估计GLSE关于误差分布的稳健性,求出误差项所对应的最大分布族,进而证明了在该区间波动情况下,误差向量对应的始终为一致最优解.     

混合系数线性模型参数的广义Liu估计

本文研究连续测量数据情况下的混合系数线性模型的参数估计问题.利用压缩估计方法给出了该模型的一类新的有偏估计-广义Liu估计,并在均方误差意义下,证明此类估计分别优于最小二乘估计、Liu估计.最后讨论参数的选取问题.     

A multinomial spline approximation scheme using spline quasi-interpolants

This paper presents a multinomial spline approximation scheme based on spline quasi-interpolants. The scheme can be considered as an extension of the usual Bernstein approximation for complex exponentials. Error estimates and numerical examples are given to show that this new scheme could produce highly accurate results.     

Minimum classification error training in example based speech and pattern recognition using sparse weight matrices

The Minimum Classification Error (MCE) criterion is a well-known criterion in pattern classification systems. The aim of MCE training is to minimize the resulting classification error when trying to classify a new data set. Usually, these classification systems use some form of statistical model to describe the data. These systems usually do not work very well when this underlying model is incorrect. Speech recognition systems traditionally use Hidden Markov Models (HMM) with Gaussian (or Gaussian mixture) probability density functions as their basic model. It is well known that these models make some assumptions that are not correct. In example based approaches, these statistical models are absent and are replaced by the pure data. The absence of statistical models has created the need for parameters to model the data space accurately. For this work, we use the MCE criterion to create a system that is able to work together with this example based approach. Moreover, we extend the locally scaled distance measure with sparse, block diagonal weight matrices resulting in a better model for the data space and avoiding the computational load caused by using full matrices. We illustrate the approach with some example experiments on databases from pattern recognition and with speech recognition.     

解不适定问题的一种多尺度方法

1 前言 数学物理反问题是应用数学领域中成长和发展最快的领域之一.反问题大多是不适定的.对于不适定问题的解法已有不少的学者进行探索和研究,Tikhonov正则化方法是一种理论上最完备而在实践上行之有效的方法(参见[5,6,7,8,13]).     

Modelling the discretization error of initial value problems using the Wishart distribution

This paper presents a new discretization error quantification method for the numerical integration of ordinary differential equations. The error is modelled by using the Wishart distribution, which enables us to capture the correlation between variables. Error quantification is achieved by solving an optimization problem under the order constraints for the covariance matrices. An algorithm for the optimization problem is also established in a slightly broader context.     

基于联系数的钢筋抗拉强度测量不确定度评定新方法

钢筋抗拉强度的不确定度包括:钢筋直径的不确定度分量,拉力的不确定度分量,检测结果重复性引入的不确定度分量,数据修约的不确定度分量等等,因此,测量不确定度是与测量数据相联系的,联系数是处理不确定性问题的一种系统数学理论,可以用联系数来表示测量不确定度,为此,提出一种基于联系数的钢筋抗拉强度测量不确定度评定的新方法.     

改进的带驻留时间隐Markov模型的前向-后向算法

对隐Maxkov模型(hidden Markov model:HMM)的状态驻留时间的概率进行了修订,给出了改进的带驻留时间隐Markov模型的结构,并在传统的隐Markov模型(traditional hidden Markov model:THMM)的基础上讨论了新模型的前向.后向变量,导出了新模型的前向-后向算法的迭代公式,同时也给出了新模型各个参数的重估公式.     

A NEW PSEUDOSPECTRAL APPROXIMATION FOR THE FOWARD-BACKWARD HEAT EQUATION

ＡＮＥＷＰＳＥＵＤＯＳＰＥＣＴＲＡＬＡＰＰＲＯＸＩＭＡＴＩＯＮＦＯＲＴＨＥＦＯＷＡＲＤ－ＢＡＣＫＷＡＲＤＨＥＡＴＥＱＵＡＴＩＯＮ￥ＹｅＸｉｎｇｄｅ（叶兴德）ＪｉａｎｇＪｉｎｓｈｅｎｇ（江金生）（Ｄｅｐｔ．ｏｆＭａｔｈ．＆Ｉｎｆｏｒ．Ｓｃｉ．，Ｈａｎｇ...     

Lower and upper bounds for characteristic functions

In the paper, several known inequalities for characteristic functions are sharpened and several new ones are derived. Proceedings of the XVII Seminar on Stability Problems for Stochastic Models, Kazan, Russia, 1995, Part III.     

VARIATION IN THE MEAN SQUARED ERROR OF ESTIMATES DUE TO FUZZINESS IN THE INFORMATION SYSTEM

In the general problem of Parametric Point Estimation the Mean Squared Error often appears as a useful measure of goodness or closeness of estimates. Nevertheless, in very rare cases an estimator with smallest Mean Squared Error exists, but Statistical Inference provides a variety of methods to find estimates. that are usually characterized by a small Mean Squared Error.When the observation of outcomes from the probabilistic information system or experiment concerning the estimation problem involves fuzzy imprecision, so that the observable events are described by means of fuzzy events on the sample space, the use of Zadeh's probabilistic definition allows us to immediately extend the Mean Squared Error.In the present paper we are going to verify that the presence of fuzziness in experimental data entails a variation in that measure of goodnesss of estimation. On the basis of the last assertion the problem of selecting the suitable sample size, in order to remove the variation in the Mean Squared Error due to fuzziness or to estimate the parameter with a specified degree of precision, will be then discussed.     

Estimation of a structural linear regression model with a known reliability ratio

In this paper, we consider the estimation of the slope parameter of a simple structural linear regression model when the reliability ratio (Fuller (1987),Measurement Error Models, Wiley, New York) is considered to be known. By making use of an orthogonal transformation of the unknown parameters, the maximum likelihood estimator of and its asymptotic distribution are derived. Likelihood ratio statistics based on the profile and on the conditional profile likelihoods are proposed. An exact marginal posterior distribution of , which is shown to be at-distribution is obtained. Results of a small Monte Carlo study are also reported.The first author acknowledges partial finantial suport from CNPq-BRASIL.