基于多元统计分析的高校课堂教学质量评价研究

本文在阐述多元统计分析理论的基础上,以本校教学评价结果为样本,对高校课堂教学质量评价进行了主成分分析及聚类分析。通过主成分分析构建综合指标,对各个原始指标的权重进行了修正,使得原始数据得到有效简化。聚类分析对各个教师进行了有效分类,使得评价结果更加科学,对课堂教学质量管理提供了有意义的信息。     

Spatial determinants of productivity: Analysis for the regions of Great Britain

This paper uses NUTS3 sub-regional data for Great Britain to analyse the determinants of spatial variations in income and productivity. We decompose the spatial variation of earnings into a productivity effect and an occupational composition effect. For the former (but not the latter) we find a robust relationship with proximity to economic mass, suggesting that doubling the population of working age proximate to an area is associated with a 3.5% increase in productivity in the area. We measure proximity by travel time, and show that effects decline steeply with time, ceasing to be important beyond approximately 80 min.     

An empirical examination of clustering and dispersion within Canadian shopping centers

In this paper, data on the internal compositions of 90 planned regional shopping centers in the five westernmost provinces in Canada are used to examine the locational pattern of stores in shopping centers, to see whether these locations are consistent with exploiting demand externalities and the physical features of the mall. The empirical relevance of so-called “rules of thumb” for locating stores in shopping centers is also assessed. We find that there is clustering of service stores near mall entrances, and clustering of comparison shopping stores near corridor intersections and on the second floor of two-story malls. Clustering tends to occur in the ladies' wear, jewellery/fashion accessory, and unisex clothing store categories, facilitating comparison shopping. Clustering of stores in the service category facilitates multipurpose shopping. A regression analysis indicates that clustering may depend upon the size, age, and type of mall in question. Overall, results are consistent with consumer transportation/shopping costs and demand externalities driving the internal store location strategy of planned regional shopping centers.     

United States digital divide: State level analysis of spatial clustering and multivariate determinants of ICT utilization

This study analyzes factors associated with availability and utilization of information and communication technologies (ICTs) for U.S. states. We construct an exploratory conceptual model of technology utilization in which eight dependent ICT availability and utilization factors are posited to be associated with fourteen independent socio-economic, demographic, innovation, social capital, and societal openness factors. Technology utilization variables are spatially analyzed to determine extent of agglomeration, and regression residuals are examined to eliminate spatial bias. Findings indicate social capital, education, societal openness, urbanization, and ethnicities are significantly associated with ICT utilization. We suggest important implications for policymakers at state and federal levels.     

基于高维聚类技术的中文关键词提取算法

关键词提取是中文信息处理技术研究中的热点和难点,基于统计信息的方法是其中一个重要分支。本文针对基于统计信息关键词提取方法准确率低的问题,提出基于高维聚类技术的中文关键词提取算法。算法通过依据小词典的快速分词、二次分词、高维聚类和关键词甄选4个步骤实现关键词的提取。理论分析和实验显示,基于高维聚类技术的中文关键词提取方法具备更好的稳定性、更高的效率和更准确的结果。     

The spatial clustering investment behavior in housing markets

Housing markets has long been studied by researchers. This paper aims to investigate the tax policy (Special Stamp Duty) impact on housing markets in spatial clustered pattern, which has scarcely been considered in previous studies. To fill in this gap, a group of second hand housing transaction records in Hong Kong is examined by spatial scan statistics and multivariate logistic regression analysis. Our methodology differs from previous studies that it is location-oriented. Our new findings point out that the recent execution of Special Stamp Duty would cause venturi effect in housing markets: the overall transaction volume would significantly move downwards and more high transaction regions would be formed. The location factor has a significant and positive relationship with the change of transaction status. Moreover, the venturi effect provides us valuable information to understand the housing investment behavior as well as the short-term feedback for government and policy makers.     

跨境贸易人民币结算对我国进口企业财务影响的实证研究

跨境贸易人民币结算政策的出台给我国经济带来的影响是巨大的,而企业作为经济组织,其财务方面是否受到影响以及受到何种程度的影响是备受关注的问题。本文通过选取28个样本企业并对其财务状况、经营成果和现金流量三个方面的20个财务指标利用SPSS统计软件采用因子分析和聚类分析相结合的方法进行实证研究,分析得出适时推出的跨境贸易人民币结算对我国进口企业财务产生了积极深远的影响。     

Hydrocarbon pollution in underwater sinkholes of the Mexican Caribbean caused by tourism and asphalt: Historical data series and cluster analysis

In this study historical data sets from 2012, 2013, 2014 and 2015 of hydrocarbon pollution in sinkholes of Quintana Roo are presented for two vacational seasons: “High” (July) and “Low” (October). Anthracene, Benzo[a]pyrene and Pyrene were found to be the main hydrocarbon contaminants and these poly aromatics hydrocarbons are strongly related to pollution by asphalt. Tourism contributes to hydrocarbon pollution in the sinkholes of Quintana Roo in two main ways: runoff of fuel from cars and, principally, from annual asphalting. Additionally, in this paper we conduct a cluster analysis to classify the pollution grade of the sinkholes. Clustering can be used to observe and interpret the cause of vacational seasonal pollution produced by tourism.     

基于线性聚类算法的轴承钢硬度涡流检测虚拟系统的设计

针对轴承钢硬度的要求，提出了一种对电磁检测中的信号进行分类处理的简便方法，得出了一种简单实用的无IR区的线性分类规则，在此基础上利用C＋＋Builder设计了轴承钢硬度虚拟检测系统。     

风险投资网络社群集聚性与可达性对成功退出的影响

从中观层面研究风险投资网络成为风险投资领域的热点。基于风险投资网络社群这一中观视角,研究风险投资网络社群集聚性、可达性及二者交互效应对机构成功退出的影响。运用CV Source数据库2005年1月1日至12月31日的样本数据,采用多元回归模型进行分析。结果表明:①风险投资机构网络社群集聚与成功退出呈倒U型关系;②可达性与成功退出呈正相关关系;③可达性正向调节社群集聚与成功退出的倒U型关系。研究结果对风险投资网络社群结构优化、提高机构投资绩效具有一定理论价值和实际意义。