基于多维数据模型的数据仓库的构建

为了更准确地把握出版行业图书等产品的发行销售和客户的购买趋势及出版行业产品供应市场的变化趋势,对在出版行业建立业务管理及辅助决策系统进行了研究.利用Microsoft提供的OLAP Services工具,根据相应的事实表和维表构建了基于MOLAP存储方式的立方体-图书销售分析数据模型,构建了基于多维数据模型的数据仓库.由原型系统研究表明,多维数据模型的数据仓库有利于出版行业市场变化分析,为企业的经营管理和决策起到重要作用.     

Alternatives to data averaging of consumer preference data

The relationship between a consumer preference data set and a corresponding sensory profile on eight cooked wheat noodles with different formulas was examined using several multivariate techniques. Individual consumer hedonic responses (100 noodle/pasta consumers) and eight appearance and texture sensory attributes were collected. The consumer preference data were treated in two different ways: mean values averaged across all consumers or principal components extracted from individual responses. The mean preference scores were submitted to both principal component stepwise regression and partial least squares regression (PLS1), whereas the summarized major preference components were subjected to canonical correlation analysis, as well as partial least squares regression (PLS2). The results suggested that in case of complex consumer data, using mean value can only capture the most manifest trends in consumer preference patterns, while studying individual responses and by further categorizing major preference patterns provide an opportunity to discover the hidden information that are masked by data averaging.     

Revisiting French tomato data: Cluster analysis with incomplete data

The analysis of French tomato data from a 2004 Sensometrics workshop is revisited. The workshop posed two questions (1) are there consumer segments in hedonic data and (2) if there are segments, can they be characterized using consumer and tomato attributes. The challenge with the hedonic data is a large amount of missing data. “Probabilistic” solutions to the latter via multiple imputation are explored. In addition to more traditional methods, polynomial models are used to answer the second question.     

用Perl编程进行因特网数据整理的方法

采用Perl编程语言对网络资源数据进行转换,使其数据格式适合S-plus软件处理.并以2001年法国赛车数据为例进行分析,结果表明Perl非常适合处理零乱的原始数据.通过识别一些特殊字符,Perl可以读取包含特殊字符的那些行,这个性质使Perl成为数据整理的有力工具.     

Integrity of small data bases in computer analysis of dietary data

The integrity of data bases to support microcomputer-based dietary analysis programs has become increasingly important to developers and users of nutritional analysis software. This paper reviews critical issues in maintaining data integrity during development of small nutritional data bases. Because a limited number of large, source data bases provides the data for smaller, special-purpose data bases, this review initially focuses on factors that affect the quality and precision of methodologies used in establishing large data bases. Issues discussed are accuracy of source data as determined by analytical methodology and imputation procedures, and methods for insuring representativeness of data. The effect of data transfer procedures on small data base integrity are discussed, including use of multiple sources and standardization of naming and coding conventions. Also reviewed are procedures for selecting reduced numbers of foods and nutrients without sacrificing accuracy of analysis, and methods currently in use for validating small data bases.     

福建烟草数据中心数据质量监控技术应用

以推广福建烟草数据中心应用及提高系统数据质量为目的,结合福建烟草商业公司数据中心平台营销、专卖、物流、人力劳资、综合计划等多个部门的业务主题建设与分析应用项目的具体需要,研究并提出了一种合适的数据中心数据质量监控方法。该方法首先在保证数据质量的前提下成功将烟草公司源系统数据抽取到省公司数据仓库中,然后在数据集市层主要利用星型关系完成数据模型设计作业,最后在BI应用中分析与展现了数据。该方法对整个数据中心的高效应用起到了至关重要的作用。      

基于大数据视角的肉类食品安全抽检数据分析

目的 基于大数据视角分析肉类食品安全抽检数据。方法 在大数据技术支持下以国家肉类食品抽检监测数据源为基础, 通过Python语言编程设计分类与预测实验, 并利用数据挖掘预测集实验结果与真实食品检验结果进行对比研究, 以验证该方法的可行性。结果 基于决策树+典型相关系数和二次指数平滑法相结合的数据挖掘方法分类效果较好, 预测准确性达到98.26%。结论 通过预判不合格肉类食品的出现数量和分布情况, 可指导其安全抽检监测工作, 提高管理的效率和准确率, 有效预防肉类食品安全事故的发生。     

Statistical analysis of sensory profiling data: data reduction and generalised Procrustes analysis

The aim of the study is to show the sensory differences between 16 varieties of potatoes assessed twice by 14 judges with 30 attributes. Instead of simply performing a statistical method taking into account the variability among judges and yielding a judge consensus space, a selection step is carried out. It aims at reducing the data to keep the information relevant to the objective of the analysis. Thus the sensory attributes are selected on their discriminative ability and on the extent to which the judges agreed on their meaning. The judges are chosen according to their performance.     

Estimating correlation with multiply censored data arising from the adjustment of singly censored data

Environmental data frequently are left censored due to detection limits of laboratory assay procedures. Left censored means that some of the observations are known only to fall below a censoring point (detection limit). This presents difficulties in statistical analysis of the data. In this paper, we examine methods for estimating the correlation between variables each of which is censored at multiple points. Multiple censoring frequently arises due to adjustment of singly censored laboratory results for physical sample size. We discuss maximum likelihood (ML) estimation of the correlation and introduce a new method (cp.mle2) that, instead of using the multiply censored data directly, relies on ML estimates of the covariance of the singly censored laboratory data. We compare the ML methods with Kendall's tau-b (ck.taub) which is a modification Kendall's tau adjusted for ties, and several commonly used simple substitution methods: correlations estimated with nondetects set to the detection limit divided by 2 and correlations based on detects only (cs.det) with nondetects setto missing. The methods are compared based on simulations and real data. In the simulations, censoring levels are varied from 0 to 90%, p from -0.8 to 0.8, and v (variance of physical sample size) is set to 0 and 0.5, for a total of 550 parameter combinations with 1000 replications at each combination. We find that with increasing levels of censoring most of the correlation methods are highly biased. The simple substitution methods in general tend toward zero if singly censored and one if multiply censored. ck.taub tends toward zero. Least biased is cp.mle2, however, it has higher variance than some of the other estimators. Overall, cs.det performs the worst and cp.mle2 the best.