Partially pooled covariance matrix estimation in discriminant analysis

The Linear Discriminant Rule (LD) is theoretically justified for use in classification when the population within-groups covariance matrices are equal, something rarely known in practice. As an alternative, the Quadratic Discriminant Rule (QD) avoids assuming equal covariance matrices, but requires the estimation of a large number of parameters. Hence, the performance of QD may be poor if the training set sizes are small or moderate. In fact, simulation studies have shown that in the two-groups case LD often outperforms QD for small training sets even when the within -groups covariance matrices differ substantially. The present article shows this to be true when there are more than two groups, as well. Thus, it would seem reasonable and useful to develop a data-based method of classification that, in effect, represents a compromise between QD and LD. In this article we develop such a method based on an empirical Bayes formulation in which the within-groups covariance matrices are assumed to be outcomes of a common prior distribution whose parameters are estimated from the data. Two classification rules are developed under this framework and, through the use of extensive simulations, are compared to existing methods when the number of groups is moderate.     

Reifications of the intellectual: representations,organization and agency in revolutionary China

How did ‘intellectuals’ evolve from a class of subjects in Marxian thoughts to highly visible populations under communism? Such ‘reifications of the intellectual’ have deeply affected subjectivity, conflict and organization, but received little attention in the political sociology of communism. This essay draws on research on classifications and social boundaries to address the objective and subjective foundations of the reifications and their impact on communist rule. The intellectual is viewed as an identification formed and performed around multiple social axes (most notably family background, educational achievement, occupational history, institutional affiliation and revolutionary rank) that reflected broader patterns of communist political domination. I use the Chinese Communist movement to demonstrate that (1) interaction of political contests, ruling strategies and institutional developments turned a diversity of persons into ‘intellectuals’ who were allegedly imbued with reprehensible interests and habits linked to privileged economic classes; (2) constant competitions for power and organizational changes led to classificatory ambiguities and, in turn, allowed individuals some control over their identifications; and (3) the developments profoundly influenced identity, state and class formation. Focusing on the dynamics that produced a highly visible but fluid population of ‘intellectuals’ opens new pathways for comparative research on communism.     

The International Classification of Functioning,Disability, and Health (ICF), A New Tool for Social Workers

Abstract

The World Health Organization (WHO) ratified the International Classification of Functioning, Disability and Health (ICF) in 2001. This followed a ten year period of international development which shifted emphasis from a system concerned with the consequences of disease to a system concerned with human functionality and health. Disability advocates contributed to this system which has potential as a tool for social work classification and assessment. Familiarity with the system also brings the profession in step with the disability studies movement which promotes a social model of disability.     

SEQUENTIAL,BOTTOM‐UP VARIABLE SELECTION FOR HIGH‐DIMENSIONAL CLASSIFICATION

Most methods for variable selection work from the top down and steadily remove features until only a small number remain. They often rely on a predictive model, and there are usually significant disconnections in the sequence of methodologies that leads from the training samples to the choice of the predictor, then to variable selection, then to choice of a classifier, and finally to classification of a new data vector. In this paper we suggest a bottom‐up approach that brings the choices of variable selector and classifier closer together, by basing the variable selector directly on the classifier, removing the need to involve predictive methods in the classification decision, and enabling the direct and transparent comparison of different classifiers in a given problem. Specifically, we suggest ‘wrapper methods’, determined by classifier type, for choosing variables that minimize the classification error rate. This approach is particularly useful for exploring relationships among the variables that are chosen for the classifier. It reveals which variables have a high degree of leverage for correct classification using different classifiers; it shows which variables operate in relative isolation, and which are important mainly in conjunction with others; it permits quantification of the authority with which variables are selected; and it generally leads to a reduced number of variables for classification, in comparison with alternative approaches based on prediction.     

搜索引擎营销及其关键词选择

搜索引擎营销是基于搜索者行为的全程营销,它能实现品牌认知、网站转化和客户服务等目标。搜索引擎营销的第一步是关键词的选择,不管是通过搜索引擎优化和付费搜索引擎广告,都需要选择能够实现目标的期望关键词。为此,必须对搜索引擎营销关键词选择作理论和实证分析,探讨搜索引擎营销关键词的选择方法。     

中国现当代文学选本的分类

中国现当代文学选本的存在形式不是稳定不变的,而是随着载体、文艺观念等不断扩展,不同时期选本的分类不尽相同。可以按照功能、载体、专题、文体、时间、民族、语言、地域,把迄今为止的中国现当代文学选本分作八个类别。     

采用非线性核支持向量机并基于基因表达数据的基因选择和分类

在基于微阵列的癌症分类中,由于变量(基因表达)较多,而实验条件较少,因此特征选择和分类方法非常重要。对于疾病诊断,分类器的性能直接影响到最终结果的准确性。本文提出一种新的基因选择和分类方法,这种方法使用基于递归特征排除(RFE)的非线性核支持向量机(SVM)。实验表明本文方法比其它线性分类方法具有更好的整体表现,如线性核支持向量机和Fisher线性判别分析方法;同样本文方法也比一些非线性分类方法更好,如采用非线性核的最小二乘支持向量机(LS-SVM)。实验除了使用测试集,还使用留一校验算法(leave-one-out)用于测试分类器的泛化性能。实验采用可通过互联网获得的AML/ALL数据集和遗传性乳腺癌数据集。     

论文化产业概念与分类

产生于不同背景和条件下的国内外诸多文化产业概念与分类之间有着明显的差异,随着现代科学技术、各国文化经济实践特别是我国文化产业的不断发展,对其进一步研究有着首要的、深远的理论与实践意义.针对文化产业概念与分类的动态发展性,应坚持普遍性和特殊性相结合,体现以人为本特色,坚持发展,尊重实践,利于创新,根据我国的国情与国际竞争状况,文化产业的特征和发展规律,产业细化、融合和兴衰演变规律,进一步完善符合我国实际与未来发展的文化产业概念与分类,以科学指导和促进文化产业的发展.     

基于最小生成树的基因分类算法

随着基因芯片技术的快速发展以及其在基因表达分析等过程中的应用,产生了大量的基因表达谱数据,如何处理和分析这些数据并从中提取出有价值的生物学信息成为一个极为重要的课题,基因分类是进行基因数据处理的常用方法。本文首先利用主成分分析法(PCA)把基因的多个属性转化为少数几个综合属性,将基因表达谱数据映射成一个带权图,并将图论的最小生成树理论引入基因分类分析方法,然后设计了基于最小生成树的基因分类算法,理论分析和仿真结果表明了该算法的可行性和有效性。     

手写汉字识别粗分类方法的研究

提出了一种基于汉字结构关系的手写汉字识到粗分类的新方法。该方法以抽取手写汉字基本笔划为基础，利用汉字的结构特征进行粗分类。它不仅算法简单、分类速度快、分类能力强、分类稳定、可靠，而且在分类的同时又能把组成汉字的部件及其笔划随之而分离出来，为用句法结构方法识别手写汉字奠定了基础。这种方法成功地用于在ＰＣ／ＡＴ机上实现的成页手写体（楷书）汉字识别系统中，效果满意。