基于贝叶斯伪设计与组合样本的候选者数据库网络调查的推断研究

候选者数据库网络调查的推断问题是网络调查发展中迫切需要解决的问题.基于此,提出基于贝叶斯伪设计与组合样本的非概率抽样推断方法:将网络候选者数据库的调查样本与概率样本结合,根据贝叶斯定理推导出网络候选者数据库的调查样本单元的伪权数构造式,再利用两个样本数据共同估计总体均值,最后利用Bootstrap和Jackknife方法来计算总体均值估计的方差估计.研究结果表明:基于贝叶斯伪设计与组合样本的总体均值估计比使用Elliot方法估计的总体均值偏差更小,估计效果较好;方差估计方面,Bootstrap方差估计比Jackknife方差估计的效果好.     

随机抽样——掌握本质,准确求解

<正>推断性统计学的基本思想方法是用样本估计总体,即通过从总体中抽取一个样本,根据样本的情况去推断总体的相应情况.因此,科学、合理地选择抽样方法采集样本,直接关系到对总体推断的准确程度.在学习中了解简单随机抽样、系统抽样、分层抽样的操作方法以及它们的区别与联系是我们解决有关统计问题的一个重点内容.     

正态总体均值已知情况下方差的统计推断问题

数理统计的任务之一是利用样本提供的信息对总体作出统计推断，但在知道总体的有关信息时就应充分利用．本文讨论了正态总体均值μ已知的情况下方差σ～２的统计推断问题．     

基于超总体伪设计与组合样本的候选者数据库网络调查的推断研究

候选者数据库网络调查的推断问题是网络调查发展中迫切需要解决的问题.基于此,提出基于超总体伪设计与组合样本的非概率抽样推断方法:对网络候选者数据库的调查样本建立超总体模型来构造伪权数,并根据网络候选者数据库的调查样本和概率样本的组合样本计算总体均值的估计,最后根据超总体模型的方差估计理论推导出目标总体均值估计的方差估计式,同时采用Bootstrap与Jackknife方法来估计总体均值估计的方差,并比较不同方差估计方法的效果.研究结果表明:基于超总体伪设计与组合样本的总体均值估计效率高于仅使用概率样本的估计和仅使用网络候选者数据库的调查样本加权的估计,估计效果较好;方差估计方面,采用VM1、VM2与VM3方法计算的方差估计相比而言更好.     

非概率样本的模型推断

非概率抽样在大数据时代有广阔的应用空间,但其统计推断问题仍有待研究和发展.针对这一问题,提出利用基于模型的推断方法结合配额抽样实现非概率样本的统计推断,其思路是先设定线性回归形式的超总体模型,再利用配额样本观测数据拟合模型估计未知参数,进而利用模型对非观测单元进行预测,案例分析结果显示基于超总体模型的推断方法是解决非概率样本统计推断的有力途径,具有较大的深入研究价值.     

基于倾向得分广义线性模型的非概率抽样统计推断研究

候选者数据库网络调查下非概率抽样的统计推断问题是网络调查发展中迫切需要解决的问题.提出基于倾向得分广义线性模型的非概率抽样推断方法:将网络候选者数据库的调查样本与参考样本结合,建立Logistic、Probit、C-log-log三种广义线性模型来估计倾向得分,并对网络候选者数据库的调查样本进行倾向得分未加权比例的分组调整与倾向得分加权比例的分组调整来估计总体.研究结果表明:基于倾向得分广义线性模型的总体估计效果较好,并且使用调查权数的Logistic与C-log-log倾向得分未加权比例的分组调整估计最为稳健.     

基于样本次序统计量的总体分位数的非参数统计推断

随机总体分位数的统计推断理论与方法一直是统计学研究的重要课题.其主要原因是分位数的应用涉及众多领域,且在各领域的研究中起到举足轻重的作用.本文系统地论述了基于样本次序统计量的总体分位数的非参数统计推断的理论和方法;给出了基于样本次序统计量的总体分位数的估计方法,总体两个分位数之差的置信区间,总体容许区间的求解方法及符号检验.希望有助于读者的科研与应用.     

多源数据融合视角下非概率样本与概率样本的大量插补推断方法

随着社会的发展,概率样本无回答率越来越高,其目标变量可能存在缺失的情况.同时,大数据与网络调查的发展使得获得的样本大多数是非概率样本,如何结合这两种样本推断总体是当今时代多源数据融合领域的一个热点问题.假设存在目标变量完全缺失的概率样本和数据完整的非概率样本,提出基于非概率样本建立超总体局部多项式模型,插补概率样本缺失的目标变量,并利用插补后的概率样本估计总体,进一步证明提出估计的渐近性质.模拟和实证研究表明:与基于非概率样本的倾向得分逆加权估计相比,提出估计的绝对相对偏差,方差与均方误差更小,且与基于真实概率样本的总体估计相接近;提出总体均值估计的方差估计的绝对相对偏差与95%置信区间覆盖率也接近于基于真实概率样本的总体估计的相应指标,估计效果较好.     

一类马氏样本下假设检验问题错误概率的估计

本文考虑样本不独立情形的统计推断问题,研究基于马氏样本的最优势检验,给出了此情形的Neyman-Pearson基本引理.当样本容易足够大时,利用大偏差原理,得到了Neyman-Pearson型检验所犯两类错误概率的精确估计,它推广了经典Neyman-Pearson基本引理的相关结果.     

NA样本下含附加信息时M-泛函的经验似然推断

本文研究NA样本下不含附加信息和含附加信息时M-泛函的经验似然推断,将分块技术应用到经验似然方法中,证明M-泛函的对数经验似然比统计量的渐近分布为卡方分布,由此构造NA样本下不含附加信息和含附加信息时M-泛函的经验似然置信区间.同时,对一类M-泛函的检验问题,证明含附加信息比不含附加信息时有更高的功效,并给出数值模拟结果.     

Efficient Computation of the Joint Sample Frequency Spectra for Multiple Populations

A wide range of studies in population genetics have employed the sample frequency spectrum (SFS), a summary statistic which describes the distribution of mutant alleles at a polymorphic site in a sample of DNA sequences and provides a highly efficient dimensional reduction of large-scale population genomic variation data. Recently, there has been much interest in analyzing the joint SFS data from multiple populations to infer parameters of complex demographic histories, including variable population sizes, population split times, migration rates, admixture proportions, and so on. SFS-based inference methods require accurate computation of the expected SFS under a given demographic model. Although much methodological progress has been made, existing methods suffer from numerical instability and high computational complexity when multiple populations are involved and the sample size is large. In this article, we present new analytic formulas and algorithms that enable accurate, efficient computation of the expected joint SFS for thousands of individuals sampled from hundreds of populations related by a complex demographic model with arbitrary population size histories (including piecewise-exponential growth). Our results are implemented in a new software package called momi (MOran Models for Inference). Through an empirical study, we demonstrate our improvements to numerical stability and computational complexity.     

Using Global Optimization to Estimate Population Class Sizes

In this paper we formulate a nonlinear optimization model to estimate population class sizes based on sample information. The model is nonconvex and has several local minima corresponding to different populations that could have been the source of the sample data. We show that many if not all local solutions can be found using a new global optimization algorithm called OptQuest/NLP (OQNLP). This can be used to estimate the number of individuals in a population with unique or rarely occurring characteristics, which is useful for assessing disclosure risk. It can also be used to estimate the number of classes in a population, a problem with applications in a variety of disciplines.     

Bayesian spatial modeling of genetic population structure

Natural populations of living organisms often have complex histories consisting of phases of expansion and decline, and the migratory patterns within them may fluctuate over space and time. When parts of a population become relatively isolated, e.g., due to geographical barriers, stochastic forces reshape certain DNA characteristics of the individuals over generations such that they reflect the restricted migration and mating/reproduction patterns. Such populations are typically termed as genetically structured and they may be statistically represented in terms of several clusters between which DNA variations differ clearly from each other. When detailed knowledge of the ancestry of a natural population is lacking, the DNA characteristics of a sample of current generation individuals often provide a wealth of information in this respect. Several statistical approaches to model-based clustering of such data have been introduced, and in particular, the Bayesian approach to modeling the genetic structure of a population has attained a vivid interest among biologists. However, the possibility of utilizing spatial information from sampled individuals in the inference about genetic clusters has been incorporated into such analyses only very recently. While the standard Bayesian hierarchical modeling techniques through Markov chain Monte Carlo simulation provide flexible means for describing even subtle patterns in data, they may also result in computationally challenging procedures in practical data analysis. Here we develop a method for modeling the spatial genetic structure using a combination of analytical and stochastic methods. We achieve this by extending a novel theory of Bayesian predictive classification with the spatial information available, described here in terms of a colored Voronoi tessellation over the sample domain. Our results for real and simulated data sets illustrate well the benefits of incorporating spatial information to such an analysis.     

On an epidemiological model with nonlinear infection incidence: Local and global perspective

Though bilinear and standard incident rates have been used frequently in classical epidemic models, nonlinearity in infection incidence is considered to be an useful modification due to its proximity with various realistic situations of disease propagation among ecological populations. In the present research, we consider an epidemiological model with SIS disease in the population. Infection is assumed to propagate following nonlinear incidence. First, we perform a stability and bifurcation analysis of the system around different equilibria from a local perspective. It is observed that the infected incidence fraction (p) plays the key role in controlling the disease dynamics. Then, using an iteration scheme and comparison argument, we investigate the global stability criteria of the model system around the endemic state and infer that the disease will persist among the species in the long run. Numerical simulation study is also carried out to illustrate and augment the analytical findings.     

高校公共就业服务个人自愿供给对就业水平影响机理的实证研究

“多中心治理”框架下完善的公共就业服务体系应该是政府供给、市场供给和自愿供给的有效融合,公共就业服务自愿供给机制能否有效弥补政府供给和市场供给的相对不足。文章构建了公共就业服务个人自愿供给影响就业的逻辑回归模型,基于4所高校3336份问卷调查的实际数据进行了实证检验,实证结果表明,无论是基于个体自选择信息还是利用个体志愿者多维行为特征的整合来确定个体志愿者的身份,大学生公共就业服务自愿供给均对就业产生显著正向影响,而且不同类型的公共就业服务自愿供给对就业影响存在显著差异,提供就业信息、传授求职经验、介绍面试技巧3类自愿供给服务对就业影响最显著,而介绍简历制作技巧类自愿供给对提升就业无影响。     

Feasibility principles for Downstream Demand Inference in supply chains

Many companies are adopting strategies that enable Demand Information Sharing (DIS) between the supply chain links. Recently, a steady stream of research has identified mathematical relationships between demands and orders at any link in the supply chain. Based on these relationships and strict model assumptions, it has been suggested that the upstream member can infer the demand at the downstream member from their orders. If this is so, DIS will be of no value. In this paper, we argue that real-world modelling requires less restrictive assumptions. We present Feasibility Principles to show that it is not possible for an upstream member to accurately infer consumer demand under more realistic model assumptions. Thus, we conclude that DIS has value in supply chains. We then move our focus to the supply chain model assumptions in the papers arguing that there is value in sharing demand information. Using a simulation experiment, we show that the value of sharing demand information in terms of inventory reductions will increase under more realistic supply chain model assumptions.     

Teaching Mathematics and Science in Urban Elementary Schools

Elementary teachers from a large urban school district with a 74 percent minority student population were surveyed to assess their capacity to provide quality mathematics and science instruction. Forty-nine percent of the surveys distributed to a random sample of teachers were returned. Both strengths and barriers to capacity building were identified. Strengths included use of collaborative student work, manipulatives, informal learning environments, and alternative assessment practices; availability of calculators and computers; and high expectations for student learning. Barriers included lack of professional development, infrequent science instruction, limited calculator use, lack of planning time, inadequate resources, and the perception that science was not valued as highly as mathematics.     

Fractal properties of percolation clusters in Euclidian neural networks

The process of spike packet propagation is observed in two-dimensional recurrent networks, consisting of locally coupled neuron pools. Local population dynamics is characterized by three key parameters – probability for pool connectedness, synaptic strength and neuron refractoriness. The formation of dynamic attractors in our model, synfire chains, exhibits critical behavior, corresponding to percolation phase transition, with probability for non-zero synaptic strength values representing the critical parameter. Applying the finite-size scaling method, we infer a family of critical lines for various synaptic strengths and refractoriness values, and determine the Hausdorff–Besicovitch fractal dimension of the percolation clusters.     

Mark-recapture techniques in statistical tests for imprecise data

We aim to construct suitable tests when we have imprecise information about a sample. More specifically, we assume that we get a collection of n sets of values, each one characterizing an imprecise measurement. Each set specifies where the true sample value is (and where it is not) with full confidence, but it does not provide any additional information.Our main objectives are twofold: first we will review different kinds of tests in the literature about inferential statistics with random sets and discuss the approach that best suits our definition of imprecise data. Secondly, we will show that we can take advantage from mark and recapture techniques to improve the accuracy of our decisions. These techniques will be specially important when the population is small enough (with respect to the sample size) that recaptures are common. They also seem to be useful when resampling techniques are involved in the decision process.