样本问题

Finishing calculation of mean value, standard deviation, skewness, kurtosis of Beta distribution.(2) Fitting parameters of many kinds of typical distribution and using residual deviation to evaluate fitting precision.(3) Using Beta distribution as an agreed indication distribution applied to many kinds of practical photoelectric measurement distributions.(4) Deriving theory formula of Bayes point estimation about Beta distribution parameters and mean value and standard deviation on the condition of mean square error loss function and supposed the prior distribution is uniform distribution.(5) Generating MCMC sample from post distribution by the method of Gibbs sample algorithm. Calculating bayes point estimation from sample on the condition of mean square error loss function. Calculating confidence interval by an approximate method to complete interval estimation.

本文的主要工作有：(1)解决了Beta分布参数a和b的精确计算以及均值、标准差、偏度、峰度的计算问题；(2)拟合出10余种典型分布的Beta分布的两个参数，并且采用剩余标准差评价该Beta分布的拟合精度；(3)对多种典型的光学与光电测量系统的测量分布进行了Beta分布统示表示；(4)在假设先验分布为均匀分布前提下，得到参数a和b以及均值μ和标准差σ在均方误差损失函数下的贝叶斯点估计理论计算公式；(5)利用直接抽样的Gibbs抽样算法，从后验分布中产生MCMC样本，从样本直接计算均方误差损失函数下的贝叶斯点估计，并使用一种近似方法计算其置信区间，完成区间估计。

By combining some experts' viewpoint and the author' s research, the author discusses such general issues as the use of biased and small sample, the lack of potential variables, exaggerating of statistical diagrams, reflect on reliability and significance test, causality, and so on.

本文从心理学研究过程的内在逻辑出发，探讨了在心理研究中应用统计应该注意的问题和可能遇到的误用现象：有偏样本与小样本的使用，潜在变量的缺失，欺骗性的统计图表，量表信度与统计显著性检验的考量，事后解释的谬误，统计关系与因果关系等。

This paper proposes an algorithm named Small Sample Incremental Biased Learning Algorithm to solve three difficulties of relevance feedback in sketch retrieval, including small sample issue, asymmetry of training data and real-time requirement.

为了解决草图检索相关反馈中小样本训练、数据不对称及实时性要求这3个难点问题，提出了一种小样本增量有偏学习算法。

In POT model, it used the technique of parameter estimator stability, making up for the limits of the currently popularly-used Mean Excess Function e, and the threshold quantitatively selected by Kurtosis Method has also been achieved, and solving the problem the illustration method couldn't be applied to.

特别是，在BMM模型中充分地考虑了子区间极值一般极限分布与极值序列极限分布之间的关系，测度了受子区间长度影响的极值VaR；在POT模型中运用参数估计量稳定性法弥补了目前普遍采用的样本平均超出量函数e法的不足，针对一些图解法无法适用的问题，实现了峰度法对阈值的定量选取，并对指数回归模型法、子样本自助法、序贯法等定量法进行了分析探讨。

The main work of this thesis is simulation study on the basis of large sample properties for two estimates by Monte Carlo method with some models. We consider the problem of the choice of window width, consistent rate and some doubtful points about consistent rate with fixed sample size.

本文的主要工作是依据大样本性质，对特定模型的这两种估计用Monte Carlo方法进行了模拟试验，讨论了长度有限的样本情况下窗宽的选择问题，比较了两种估计的拟合程度，相合速度并提出了对相合速度的一些疑难点。

Fourth, the method of real-time data disposal and the generalized distributing distance rule and algorithm of samples are presented by which few can be selected from plenty of measuring data.

提出过程数据预处理办法及样本点广义距离均匀分布规则和算法，解决了从海量过程测量数据中撷取少量样本数据的问题。

And then fuzzy least square support vector machines are proposed based on support vector domain description. Data samples in the feature space are described and the smallest enclosing hypersphere is obtained. The fuzzy membership value to each sample point is determined according to the distance of each sample from the center of the hypersphere, which can reduce the effect of outliers.

该方法先对样本进行数据域描述得到一个包含该组数据的最小半径的超球，再根据特征空间中样本与超球球心的距离确定它们的隶属度，减少了奇异点的影响；把所要求解的约束凸二次优化问题转化为正定线性方程组，并采用快速Cholesky分解的方法求解该方程组。

Based on the results of the iterated logarithm law and the uniform modulus of continuity for Brownian motion, Orey and Taylor(1972) discussed the multifractal decomposition of one dimension white noise.

但有关多维白噪音的重分形分解问题，即多指标Wiener过程样本轨道的重分形分析性质至今仍未见有讨论，本文旨在讨论多指标Wiener过程样本轨道的重分形分析性质。

In chapter 2 we obtain existence and joint continuity of the local time foradditive Lévy process under some conditions and discuss the upper bounds andlower bounds for the Holder laws of local times processes,the probability of somerare events ralated to small fluctuation of the sample functions and sojourn times insmall balls.Finally some liminf's results about the local and uniform oscillationsof additive stable process are acquired.

第二章给出了可加Lévy过程的局部时存在和联合连续的条件，讨论了其局部时过程的〓律上界和下界，以及可加Lévy stable过程跟样本函数局部振动相关事件的概率和停留在小球上的逗留时问题，同时对可加Lévy stable过程的样本轨道的局部和一致振动证明了一些liminf型结果。

In the 1990s, a more complete theoretical system - the achievement of statistical learning theory and neural networks, and other than because of the emerging machine learning methods of the difficulties encountered some important, such as how to determine the issue of network structure, overfitting and underfitting, the local minimum points ect, makes the rapid development and improvement of SVM in resolving the small sample, non-linear and high-dimensional pattern recognition problems in the performance of many unique advantages and can function to promote the use of fitted function and other machines learning problems.

上世纪90年代，一个较完善的理论体系--统计学习理论( Statistical Learning Theory，简称SLT)的实现和由于神经网络等较新兴的机器学习方法的研究遇到一些重要的困难，比如如何确定网络结构的问题、过学习与欠学习问题、局部极小点问题等，使得SVM迅速发展和完善，在解决小样本、非线性及高维模式识别问题中表现出许多特有的优势，并能够推广应用到函数拟合等其他机器学习问题中。

And Pharaoh spoke to Joseph, saying, Your father and your brothers have come to you.

47：5 法老对约瑟说，你父亲和你弟兄们到你这里来了。

Additionally, the approximate flattening of surface strip using lines linking midpoints on perpendicular lines between geodesic curves and the unconditional extreme value method are discussed.

提出了用测地线方程、曲面上两点间短程线来计算膜结构曲面测地线的方法，同时，采用测地线间垂线的中点连线和用无约束极值法进行空间条状曲面近似展开的分析。