聚类方法在高等学校绩效评价中的应用

高等学校都试图构建绩效评价体系,增强本校的竞争力和综合实力。本文以绩效评价、平衡计分卡等理论为指导．借鉴国内外高校绩效评价的经验,建立了高等学校绩效评价指标体系。指标体系综合考虑高等学校的财务绩效、办学效果、内部管理、学习与成长等4个方面。并且通过建立的指标体系,应用聚类挖掘方法,分别横向和纵向分析了教育部直属的72所高校2002—2006年的绩效评价指标数据,指出高等学校绩效评价分析系统中存在的问题,评价结果的信息主要用于帮助各级管理人员迅速获得高校运行的准确信息,在没有限定类别的情况下,运用数据挖掘的聚类方法把评价指标相似的学校划分到一个类中。对高等学校进行聚类评价,为分析使用提供依据,以期构建一个有利于高等学校绩效评价的决策支持环境。     

最优货币区理论在中国的应用研究——基于聚类分析方法

自蒙代尔提出最优货币区理论后,就引起了学术界广泛关注。他指出,最优货币区的经济单位是区域,而不是主权。最优货币区可以跨国界,也可以在一国以内。本文创新地把聚类分析方法用于对中国最优货币区的分类问题,说明在中国内部确实存在几个最优货币区,且指出每个区域的地区构成。最后,提出由于中国内部存在多个货币区域,统一的货币政策必然带来区域效应,中国必须加快最优货币区进程。     

Disruptive technology roadmaps

Disruptive technologies create growth in the industries they penetrate or create entirely new industries through the introduction of products and services that are dramatically cheaper, better, and more convenient. These disruptive technologies often disrupt workforce participation by allowing technologically unsophisticated individuals to enter and become competitive in the industrial workforce. Disruptive technologies offer a revolutionary change in the conduct of processes or operations.Disruptive technologies can evolve from the confluence of seemingly diverse technologies or can be a result of an entirely new technological investigation. Existing planning processes are notoriously poor in identifying the mix of sometimes highly disparate technologies required to address the multiple performance objectives of a particular niche in the market. For a number of reasons, especially the inability to look beyond short-term profitability, and the risk/return tradeoff of longer term projects, it is suggested that current strategic planning and management processes promote sustaining technologies at the expense of disruptive technologies.We propose a systematic approach to identify disruptive technologies that is realistic and operable and takes advantage of the text mining literature. This literature-based discovery process is especially useful in identifying potential disruptive technologies that may require the input from many diverse technological and management areas. We believe that this process holds great potential for identifying projects with a higher probability of downstream success. Further, we suggest a process to take the identified potential disruptive technology from the “idea stage” through to the development of a potentially feasible product for the market. This second stage makes use of workshops and roadmapping to codify the ideas of technological and management experts, who were identified in the literature-based discovery stage. Our goal is to describe and explain the pragmatic steps suggested by our innovative and practical process.The proposed process could identify technologies whose eventual development and application to specific problems would generate innovative products. The goal is to isolate technologies that have the potential to redefine an industry, or alternatively, have the potential to create an entirely new industrial setting. Use the text-mining component of literature-based discovery to identify both the technical disciplines that are likely candidates for disruptive technological products, and experts in these critical technical and managerial disciplines. While we know that this is but one way to investigate nascent disruptive technologies we feel it is imperative that the representatives of these potentially critical technical disciplines are included in the roadmap development process, either as implementers or as consultants.Every firm is looking for “the next great thing”. Literature-based discovery offers a starting point for identifying at least a portion of the major contributory technical and managerial disciplines necessary for potential disruptive technologies and discontinuous innovations. Combining literature-based discovery with a practical workshop/roadmap process dramatically enhances the likelihood of success.     

农村法治教育有效途径聚类分析

法治教育途径是农村法治教育活动的基本构成要素,也是影响农村法治教育质量与效果的关键要素之一。通过实地调查并运用聚类分析法对取得的相关数据进行分析,可以发现当前我国农村法治教育主要途径的实际有效性与理论上不一致。这启示我们必须加强对农村法治教育的组织引导,根据农村经济社会发展的现状、农民的认知水平与心理认识规律,选择合适的法治教育途径,并综合运用各种途径,使其相互配合,取长补短,才能确保农村法治教育取得应有的效果。     

基于聚类分析的自动文摘中主题区域划分方法研究

针对在海量数据中,如何有效地自动获取文摘以提高检索效率的问题,本文提出了一种自动文摘中主题区域划分的方法。该方法对文章段落向量模型进行聚类分析,得到文章的主题结构。这种方法适用于各种风格的文体,能有效解决文章主题分布自由的问题,准确地划分出文章主题区域。     

中国双向直接投资与企业创新的耦合研究——基于省级层面数据的实证分析

借鉴物理学中的耦合度函数,提出双向直接投资与企业创新的耦合协调度模型及其指标体系,而后利用中国28个省域2004—2013年的相关数据,分别计算出各省域双向直接投资与企业创新耦合协调系统的耦合度、耦合协调度,以及双向直接投资子系统与企业创新子系统的发展水平,通过比较两个子系统的发展水平将各省域分为同步型、双向直接投资驱动型与企业创新驱动型。同时,分析中国省域双向直接投资与企业创新耦合协调系统的耦合度时序变化趋势。然后,对其耦合协调度进行聚类分析,划分为协调、磨合、阻抑三个阶段。研究发现：从国家层面看,整体耦合度不高,但呈现缓慢增长趋势;从省域层面看,存在明显地区差异且各地区耦合协调阶段不同。     

本次经济危机主要大宗商品期货价格波动性研究

本文在分析本次次债危机以来主要大宗商品价格变动情况的基础上,建立了ARMA和E-GARCH模型及ARMA和TARCH模型来描述本轮经济周期中石油、铜、铝、黄金、大豆和玉米等大宗商品的期货价格收益序列的波动性特征。自相关分析发现不同商品期货市场的有效性略有差异,原油和铜市场更为有效;ARMA和非对称GARCH模型表明,主要商品收益波动均具有积聚效应,原油和铝收益波动具有杠杆效应,坏消息对原油收益波动的冲击大于好消息,好消息对铝收益波动的冲击大于坏消息。     

中关村高新技术企业合作申请专利网络集聚效应分析

根据中关村管委会提供的“一区十六园”和11个技术领域合作申请专利数据,构建高新技术企业合作申请专利网络。以模块度作为描述性指标,采用Louvain聚合算法分析社团结构特征及演化趋势。最后,具体分析4个时间截面的五大社团结构,从而揭示网络社团结构对企业合作申请专利的推动作用。     

Comparisons of the structure and infrastructure of Chinese and Indian Science and Technology

A comparison was made of the research output literatures of India and China. Both bibliometric and computational linguistics approaches were used in the comparison. China has rapidly outpaced India in both volume and citation performance of publications. China's rapid publication growth rate over the past two decades is continuing, while India's is re-starting after a relatively dormant period of almost two decades.     

提升新疆棉花综合生产能力的新思路：南疆地区棉花主体功能区建设

本文依据综合生产能力的投入产出理论建立指标体系,运用聚类分析法把新疆南疆棉花主产区的42个县市划分为四类区域,并对各个区域农业综合生产能力进行排序、归类,即I类区域为棉花综合生产能力最强的区域,依次类推。分析结果显示,I类区域有3个县市,占7．1％;II类区域有12个县市,占28．6％;III类区域有18个县市,占42．9％;IV类区域有9个县市,占21．4％。该研究旨在加快新疆南疆地区棉花主体功能区建设,即对优先发展棉区、重点发展棉区和限制发展棉区的县市“对症下药”,并进一步建立生态补偿机制,促进新疆棉花产业新发展。