短记忆性时间序列的建模分析及预测

AR模型作为时间序列模型的一种,由于其参数估计和定阶简单而广泛用于系统辨识.在多维AR序列的最小二乘建模的基础上,结合Kalman滤波算法,推导了应用Kalman滤波技术的多维AR序列参数估计方法.该算法无需保存历史数据,可对AR模型的估计参数进行实时的修正.在确定AR模型阶数时,提出了两步F检验法.选取上证某A股收益序列作为样本,利用时间序列相关性分析对该算法的有效性进行验证;对时间序列的RMSE和MAD指标进行比较,结果表明该算法大大减少了建模过程中的计算工作量,并具有较好的预测性.     

基于EMD-AR和关联维数的轧机主传动系统故障诊断研究

为提高轧机主传动系统故障诊断的精度,提出了一种基于EMD-AR模型和关联维数的故障特征提取算法.该方法采用小波滤波和EMD对振动信号进行去噪和平稳化预处理,再对包含主要故障信息的本征模函数(IMF)分别建立AR模型进行降维,最后通过时延相图法重构AR模型参数的相空间,并计算其关联维数.实验分析表明:该算法不仅能够深刻、...     

面向识别未知旋转的3维网格模型的矢量型球面卷积网络

目的 3维目标分类是视觉领域的一个基本问题,3维目标的旋转变化给分类带来极大挑战。同时不规则3维网格模型难以运用传统2维卷积网络提取特征。针对这两个问题,提出一种基于矢量型球面卷积网络的分类方法,用于识别未知旋转的3维网格模型。方法 使用矢量型神经元作为网络的基础神经元,并提出一种新型矢量层间的卷积方式。首先,将3维模型规范化并映射到单位球上,获取球面的信号表示;然后,使用矢量型分类网络和重建网络学习等变的3维模型特征;最后,使用分类网络完成3维模型分类。结果 经过消融实验对比,使用本文提出的球面卷积模块和矢量卷积层,并在训练时加入重建模块。对原本未旋转（no rotation,NR）数据集进行任意旋转（arbitrary rotation,AR）,并设定NR/AR,AR/AR,NR/NR共3种训练/测试策略的分类任务,其中NR/AR任务衡量模型识别未知旋转的能力。在刚性数据集ModelNet40上,相比基于球面卷积网络（spherical convolutional neural network,SCNN）的分类方法,在3种任务上分别提高了7.7%,1.8%,3.1%。为验证本文方法...     

基于高阶AR模型的网络异常检测

介绍基于1阶AR模型的检测方法,用高阶AR模型代替1阶AR模型以增加其精度。鉴于高阶AR模型最优参数的确定是个NP完全问题,利用遗传算法搜索其参数的近似最优解,通过实验验证高阶AR模型对网络异常检测的有效性和精确性。     

腕力传感器多维时序建模

腕力传感器有六路输出信号,表示三维空间的力和力矩信息。本文采用多维时间序列法,首次建立腕力传感器多维AR模型,其结果与实验数据吻合,全面反映动态响应过程和各输出量之间的联系。     

基于小波变换和AR模型的脑电信号注意力状态识别

针对神经反馈训练研究中注意力状态识别精度不高的特点,建立了一种基于离散小波变换和AR模型的方法用于提高注意力识别的精度。利用db N小波函数对脑电数据样本进行多层小波分解,提取小波变换系数的三种统计特征;同时采用Brug算法得到AR模型系数;用这两种参数相结合后使用支持向量机进行识别。与只采用小波变换或AR模型等特征提取方法相比,本文方法有更高的识别率。     

超宽带多径信道的频域建模与仿真

对于超宽带室内多径信道,由于信号传播的成簇特性,一般常采用较多参数描述的修正的S-V模型,而近年来频域自回归（AR）模型以其独有的特点得到大家的重视。针对频域AR模型,着重讨论了在超宽带室内多径信道的计算机仿真中,如何利用零极点分布进行频域AR建模的方法,并结合实例给出了仿真结果,说明了频域AR模型只需要很少的模型参数就可得到一个比较精确的信道描述,是超宽带信道建模比较理想的选择。     

离散序列AR模型定阶方法研究

介绍了AR模型及其参数计算方法,结合具体实例,用Gram-Schmidt正交法、奇异值分解法、LD递推法,编写程序,确定AR模型有效阶数,并进行比较分析。结果表明:Gram-Schmidt正交法所得模型的阶数、计算时间、精确度都较为适中,适合在AR模型的研究中推广普及。     

井下UHF宽带传播多径信道的频域AR建模

根据信道脉冲响应波形,文章对井下UHF宽带无线电波多径信道进行了频域自回归(AR)建模与仿真,证明了频域AR模型在井下多径信道应用的可行性,这为井下宽带乃至超宽带无线通信的信道模型计算机仿真研究提供了一种新思路。仿真结果表明,频域AR模型只需要很少的模型参数就可给出一个比较精确的信道描述,是信道建模比较理想的选择。     

基于Matlab的AR模型参数估计

AR(自回归)模型是时域信号分析中的一种常见模型。基于Matlab用时间序列的最小二乘估计和FPE、AIC准则对AR(n)模型进行参数估计的传统方法为利用多次复杂的矩阵转换求解。在传统理论的基础上,笔者利用Matlab自有函数对AR(n)模型进行参数估计,确定出AR(n)模型的最优阶次,极大简化了原有程序,分析简单易懂,增强了参数估计结果的可操作性。     

INTRODUCTION TO THE SPECIAL ISSUE ON CASE-BASED REASONING IN THE HEALTH SCIENCES

There has been an explosion of interest in health sciences applications of case-based reasoning (CBR), not only in the traditional CBR in medicine domain, but also in bioinformatics, enabling home health-care technologies, CBR integration, and synergies between CBR and knowledge discovery. This special issue features the best papers from the third workshop on CBR in the health sciences, held at ICCBR-05 in Madrid. It is the third in a series of exciting workshops, the first two of which were held at ICCBR-03, in Trondheim, Norway, and at ECCBR-04, in Madrid, Spain. The nine high-quality papers introduced here represent the research and experience of twenty-two authors working in eight different countries on a wide range of problems and projects. These papers illustrate some of the major trends of current research in CBR in the health sciences, and represent overall an excellent sample of the most recent advances of CBR in the health sciences.     

The application of IT in organisations: Some trends and issues

Abstract

This paper describes some trends and issues in the application of information technology in organisations. Technology trends are outlined in the areas of computing power, telecommunications, networking, software and standards. Some trends in the applications of IT are described with respect to developments in information systems and office automation. Some differences in the likely role of IT in the industrialised and developing countries are discussed. The second half of the paper deals with issues in the management of IT applications in organisations, and issues are identified as being political, organisational and social in nature in addition to technical. A social systems framework is proposed for the analysis of these issues. Some specific IT issues are then discussed involving alternative stakeholder perspectives, the analysis of decision making processes, the process of consultation and communication and the organisational impacts of IT applications.     

The Old-Age Healthy Dependency Ratio in Europe

The aim of this study is to answer the question of whether improvements in the health of the elderly in European countries could compensate for population ageing on the supply side of the labour market. We propose a state-of-health-specific (additive) decomposition of the old-age dependency ratio into an old-age healthy dependency ratio and an old-age unhealthy dependency ratio in order to participate in a discussion of the significance of changes in population health to compensate for the ageing of the labour force. Applying the proposed indicators to the Eurostat's population projection for the years 2010-2050, and assuming there will be equal improvements in life expectancy and healthy life expectancy at birth, we discuss various scenarios concerning future of the European labour force. While improvements in population health are anticipated during the years 2010-2050, the growth in the number of elderly people in Europe may be expected to lead to a rise in both healthy and unhealthy dependency ratios. The healthy dependency ratio is, however, projected to make up the greater part of the old-age dependency ratio. In the European countries in 2006, the value of the old-age dependency ratio was 25. But in the year 2050, with a positive migration balance over the years 2010-2050, there would be 18 elderly people in poor health plus 34 in good health per 100 people in the current working age range of 15-64. In the scenarios developed in this study, we demonstrate that improvements in health and progress in preventing disability will not, by themselves, compensate for the ageing of the workforce. However, coupled with a positive migration balance, at the level and with the age structure assumed in the Eurostat's population projections, these developments could ease the effect of population ageing on the supply side of the European labour market.     

Access to information in Europe

Developments in telecommunications, together with the availability of large collections of bibliographic references in computer-readable form, have produced a great change in access to scientific and technical information. Although the major information systems are in the USA, there is rapid development in Europe under the stimulus of the Commission of the European Communities' action plans in this area. Progress in the provision of information systems in Europe is described, and the problems that are still under consideration are discussed.     

生物传感器检测牛奶病原微生物的研究进展

牛奶中病原微生物的含量是衡量牛奶质量的重要指标之一,生物传感器在牛奶病原微生物检测方面有较好的优势.介绍了生物传感器的组成和原理,并详细综述了近几年电化学生物传感器和质量敏感型生物传感器在牛奶中病原微生物的定量检测中的研究现状,探讨了未来的发展趋势.     

The scope of digital image media in art education

In this case study of forms 7–9 in a Swedish school, the subject conceptions of and teaching practices in art subjects of schoolteachers and pupils are studied, in particular with regard to digital media. How the core content of a subject is conceived is compared to the importance of digital media in the teaching practice. For three years a class was followed whose teachers taught the pupils in both the art subject and in a local optional subject called Media. The method of triangulation involving participant observations, focus talks, interviews and questionnaires was employed in order to study the role of digital media in different subjects and their actual use by pupils. The results show that the school subject paradigm in the subject of art is chiefly connected to image production for developing aesthetic-practical skills. Traditional manual production is encouraged, thereby contributing to the limited use of digital media in the subject of art. Frame factors such as time and material also contribute to the limited use of digital media in the subject of art. Unlike the subject of art, the optional subject of media is conceived of more as a communication subject, and digital technology for image production is encouraged. In general, the use of digital media in the school is mainly guaranteed by the school's media plan. The role of the art subject is not prominent in this plan, but its teachers are active in implementing digital media in other contexts than within the framework of the art subject. In this study the representatives of the subject of art hardly used digital image processing at all in their teaching, but did so to a great extent in the optional subject of media. This made it clear that digital media in the subject of art in this school are not regarded as a prioritised media-specific competence. Offering and being able to provide a larger repertoire of digital media in the teaching of art could contribute to a change of the subject paradigm in a more clearly communicative than aesthetic-practical direction.     

基于Dijkstra算法的簇头多跳WSN监测系统

针对传统LEACH协议及其改进算法的缺点,在最小路径Dijkstra算法的基础上,将Dijkstra算法应用到LEACH协议中,旨在提高大范围监测环境中基于LEACH协议的无线网络性能。详细论述了基于Dijkstra算法的稳定数据通信阶段工作过程,针对实际监测环境采用Matlab仿真模型对改进的算法进行了实验验证。两组实验结果表明,在小范围监测环境中,基于传统LEACH协议的网络与采用Dijkstra算法完成簇间通信的LEACH协议的网络性能相似,但是随着监测范围的增大,基于改进后算法的网络性能要远远优于传统LEACH的网络性能。     

Adaptive solutions to the mutual exclusion problem

Summary Algorithms for mutual exclusion that adapt to the current degree of contention are developed. Afilter and a leader election algorithm form the basic building blocks. The algorithms achieve system response times that are independent of the total number of processes and governed instead by the current degree of contention. The final algorithm achieves a constant amortized system response time. Manhoi Choy was born in 1967 in Hong Kong. He received his B.Sc. in Electrical and Electronic Engineerings from the University of Hong Kong in 1989, and his M.Sc. in Computer Science from the University of California at Santa Barbara in 1991. Currently, he is working on his Ph.D. in Computer Science at the University of California at Santa Barbara. His research interests are in the areas of parallel and distributed systems, and distributed algorithms. Ambuj K. Singh is an Assistant Professor in the Department of Computer Science at the University of California, Santa Barbara. He received a Ph.D. in Computer Science from the University of Texas at Austin in 1989, an M.S. in Computer Science from Iowa State University in 1984, and a B.Tech. from the Indian Institute of Technology at Kharagpur in 1982. His research interests are in the areas of adaptive resource allocation, concurrent program development, and distributed shared memory.A preliminary version of the paper appeared in the 12th Annual ACM Symposium on Principles of Distributed ComputingWork supported in part by NSF grants CCR-9008628 and CCR-9223094     

Food security problems in sub-Saharan Africa: Operations Research as a tool of analysis

For many years, the author has been involved in teaching and research in the use of Operations Research as a tool of analysis to study food security problems in sub‐Saharan Africa, in particular, grass root problems of poor farmers. The paper presents an introduction to the way Operations Research methods have been applied in case studies and research projects, and discusses in retrospect the author's views on the strengths and limitations of the application of Operations Research. The paper has in particular been written for people who are not familiar with applications of Operations Research in agriculture, and are interested to learn about its potential usefulness in practice. The retrospective part is largely based on food security studies in e.g. Tanzania, Burkina Faso, Bénin, Togo and Eritrea and on participation in several interdisciplinary research programmes in Africa.