促进网络投资的互联网骨干网互联结算规制改革研究

本文研究结果表明,不管是上游市场处于垄断还是存在多家骨干网竞争情形,无互联结算规制下的网络投资量大于互联结算规制下的网络投资量,并且此时的社会福利也大于互联结算规制下的社会福利。因此,为促进骨干网市场的网络投资,应充分发挥市场机制在互联结算中的调节作用,同时要在上游市场引入竞争机制,让新的竞争者参与竞争。     

基于Copula理论的网络社交行为轨迹与金融信用相关性研究

"互联网(+)"以及大数据技术为构建互通互联的信用评价体系提供了新的契机,可解决长期以来信用指标片面化、离散化、静止化的缺陷.考虑到互联网数据多种类、多层次、多结构的特点,传统的线性回归已不适用.针对淘宝店铺和对应的新浪微博社交账号的真实数据,通过建立适当的Copula模型,对网络社交行为与金融信用之间的相关性进行探索性实证分析,证实了其相关性的成立,同时检验了Copula模型的优越性,为互联网大数据纳入信用评估体系提供了方法与论证,对促进互联网社交行为数据的价值开发、互联网金融的发展以及我国诚信体系的建设有一定的参考价值.     

复杂网络上的病毒免疫策略研究

为了给预防病毒传播提供指导意见并且更好地对病毒传播行为进行预测和控制,主要研究了几种经典复杂网络中病毒传播的模型,并对几种复杂网络病毒免疫的模型特点进行了分析,通过对这些病毒免疫模型在多局域加权网络中应用不足的分析,对多局域加权网络的病毒免疫策略进行了相应的研究.     

基于网络效用最大化的多路径网络拥塞控制研究

随着互联网的飞速发展,互联网用户和应用都在快速增长,对于网络服务质量的要求也越来越高,拥塞控制已经成为一个十分重要的问题.基于网络效用最大化的模型研究了多路径网络的拥塞控制问题,将此问题建模为基于用户发送速率的凸规划问题,基于对偶分解得到了分布式的求解方法,仿真结果验证了算法的收敛性.研究内容对多路径的TCP开发具有一定的参考意义.     

直觉模糊数加权高度排序法北大核心CSCD

利用直觉模糊数的高度与左、右扩展高度差这两者之间的差异程度,文章提出了一种新的直觉模糊数加权高度排序法,证明该排序法满足模糊数排序法7个合理性公理中的6个公理,并对其与加权均值面积排序法的关系进行了比较分析.分析说明,所提出的加权高度排序法优于加权均值面积排序法.     

一个新的模糊聚类有效性指标

提出了一个判别模糊聚类中聚类数有效性的新指标.首先利用FCM算法对数据集进行模糊聚类,通过隶属度矩阵和聚类中心构建加权二分网络.然后通过改进加权二分网络的模函数,定义一个新的聚类有效性指标.为了检验该有效性指标的性能,选取了三个常见的有效性指标在十五个数据集上进行了对比.实验结果表明,该有效性指标具有较好的性能.     

直觉模糊数加权高度排序法

利用直觉模糊数的高度与左、右扩展高度差这两者之间的差异程度,文章提出了一种新的直觉模糊数加权高度排序法,证明该排序法满足模糊数排序法7个合理性公理中的6个公理,并对其与加权均值面积排序法的关系进行了比较分析.分析说明,所提出的加权高度排序法优于加权均值面积排序法.     

直觉模糊熵与加权决策算法

直觉模糊熵是直觉模糊集理论中的一个重要概念,反映了直觉模糊集的模糊程度和不确定程度.首先给出一种新的直觉模糊熵,并运用到多属性直觉模糊决策问题中.决策时根据直觉模糊熵计算属性权重,再综合决策者的偏好对各属性权重进行修正,然后使用直觉模糊集结算子和得分函数对方案进行排序,从而获得最优方案.     

2010上海世博会影响力的定量评估——影响力的点刺激连带效应定量评估模型

利用互联网数据,在合理假设的前提下,给出了上海世博会对上海旅游业影响的定量评估.利用线性加权综合分析法建立综合评价指标的线性模型,并从两个不同的角度,分别得出上海世博会对上海旅游业的综合评价指标,同时建立灰色系统预测模型,利用2006至2009年的评价指标,对2010年的评价指标进行预测.通过实际的增长值与预测增长值的...     

一类中立型时滞互联系统的记忆反馈控制及稳定性条件

针对一类中立型时滞互联系统,当滞后时间对系统影响较小时,研究其记忆反馈控制问题.设计记忆反馈控制器,给出新的李雅普诺夫函数,结合自由加权矩阵及牛顿-莱布公式对系统稳定性进行分析,并应用线性矩阵不等式(LMI)给出中立时滞互联系统稳定的充分条件.数值仿真验证了所提控制方法的可行性.     

Exponential stability and periodicity of cellular neural networks with time delay

This paper investigates the problem of exponential stability and periodicity for a class of delayed cellular neural networks (DCNN’s). By dividing the network state variables into subgroups according to the characters of the neural networks, some sufficient conditions for exponential stability and periodicity are derived via constructing Lyapunov functional. Those conditions suitable are associated with some initial value and are represented by some blocks of the interconnection matrix. Two examples are discussed to illustrate the main results.     

复杂金融网络收益率加权和的不变性研究

对包含两个子网络的复杂金融网络进行了分析,研究了网络中各节点收益率与资金流通量之间的关系.通过建立数学模型,证明了网络中各节点收益率加权和的不变性.     

On optimal management of resources in distributed networks

In a network of processors, a distributed operating system must handle the management of shared resources. In this paper, it is shown how to solve this problem in using the model previously introduced in [1]. This model (interconnection of N Markov chains each representing locally a distributed process) allows us to prove the good functioning properties for some distributed problems such as the mutual exclusion problem and the deadlock problem, We also prove that fairness is a basic notion for setting the model’s parameters and obtain an optimal working of the network.     

Interest point detection in images using complex network analysis

This paper presents a new application of complex network theory and tools to digital image analysis and computer vision problems in order to detect interest points in digital images. We associate a weighted geometrical and fast computable complex network to each image and then we propose two different methods to locate these feature points based on both local and global (spectral) centrality measures of the corresponding network.     

Tradeoffs between quality-of-control and quality-of-service in large-scale nonlinear networked control systems

In this paper we study input-to-state stability (ISS) of large-scale networked control systems (NCSs) in which sensors, controllers and actuators are connected via multiple (local) communication networks which operate asynchronously and independently of each other. We model the large-scale NCS as an interconnection of hybrid subsystems, and establish rather natural conditions which guarantee that all subsystems are ISS, and have an associated ISS Lyapunov function. An ISS Lyapunov function for the overall system is constructed based on the ISS Lyapunov functions of the subsystems and the interconnection gains. The control performance, or “quality-of-control”, of the overall system is then viewed in terms of the convergence rate and ISS gain of the associated ISS Lyapunov function. Additionally, the “quality-of-service” of the communication networks is viewed in terms of the maximum allowable transmission interval (MATI) and the maximum allowable delay (MAD) of the network, and we show that the allowable quality-of-service of the communication networks is constrained by the required ISS gains and convergence rate of the hybrid subsystem corresponding to that network. Our results show that the quality-of-control of the overall system can be improved (or degraded) by improving (or relaxing) the quality-of-service of the communication networks. Alternatively, when relaxing the quality-of-service of one communication network, we can retain the quality-of-control of the overall system by improving the quality-of-service of one or more of the other communication networks. Our general framework will formally show these intuitive and insightful tradeoffs.     

Interconnections and Initial Conditions of Linear Systems With First-Order Representations

This paper considers the initial value problem of an interconnection composed of linear systems described by the first-order differential/algebraic equations (DAEs). An initial condition of the system variable for which the DAE has a solution is called admissible. For the interconnected system, we formulate the invariance of the admissible initial condition sets (AICSs) of the sub-systems under interconnection. Namely, the AICSs are said to be invariant if they remain unchanged even when additional constraints due to interconnection are imposed on the system variables. It is shown that the feedback and regular feedback structures of the interconnection guarantee the invariance of the AICSs in the senses of impulsive-smooth distributions and smooth distributions, respectively. The results in this paper justify the use of a feedback controller in the control system design.     

Adaptive cluster synchronization in networks with time-varying and distributed coupling delays

This paper studies the adaptive cluster synchronization of a generalized linearly coupled network with time-varying delay and distributed delays. This network includes nonidentical nodes displaying different local dynamical behaviors, while for each cluster of that network the internal dynamics is uniform (such as chaotic, periodic, or stable behavior). In particular, the generalized coupling matrix of this network can be asymmetric and weighted. Two different adaptive laws of time-varying coupling strength and a linear feedback control are designed to achieve the cluster synchronization of this network. Some sufficient conditions to ensure the cluster synchronization are obtained by using the invariant principle of functional differential equations and linear matrix inequality (LMI). Numerical simulations verify the efficiency of our proposed adaptive control method.     

Threshold analysis of the susceptible-infected-susceptible model on overlay networks

In this paper, we study epidemic spreading on overlay networks in which n multiple sets of links interconnect among the same nodes. By using the microscopic Markov-chain approximation (MMA) approach, we establish the conditions of epidemic outbreak for two kinds of spreading mechanisms in such an overlay network: the concatenation case and the switching case. When a uniform infection rate is set in all the subnetworks, we find the epidemic threshold for the switching case is just n times as large as that of concatenation case. We also find that the overlay network with a uniform infection rate can be considered as an equivalent (in the sense of epidemic dynamics and epidemic threshold) weighted network. To be specific, the concatenation case corresponds to the integer weighted network, while the switching case corresponds to the fractional weighted network. Interestingly, the time-varying unweighted network can be mapped into the static weighted network. Our analytic results exhibit good agreement with numerical simulations.     

基于改进加权Poisson CUSUM控制图的发病率监控研究

现代社会医疗质量管理的重要性不言而喻,如何科学有效地开展发病率监控,具有重要的研究意义和实用价值。本文主要研究加权Poisson CUSUM控制图的改进及其在发病率监控上的应用。通过引入"标准化"处理思想,即将对数似然比统计量除以对应样本容量,本文提出"标准化"对数似然比加权Poisson CUSUM控制图,用以提高对发病率的监控水平。通过设定合适的人口模型,模拟计算验证了本文设计的模型能够有效提高监控效果,且对于不同人口模型和不同加权函数均有显著优势。最后,基于美国新墨西哥州男性甲状腺癌患病数据开展的实证检验也印证了以上结论。