基于生物脑网络连接特性的高效神经网络算法

人工神经网络发展至今,已经在计算机视觉、类脑智能等方面得到广泛应用.在过去几十年中,人们对神经网络的研究注重追求更高的准确率,从而忽略了对网络计算成本的控制.而人脑作为高效且节能的网络,其对人工智能的发展起到了重要启示作用.如何仿真生物脑网络的连接特性,建立超低能耗的人工神经网络模型实现基本相同的目标识别正确率成为当前研究的热点.为建立低能耗的人工神经网络模型,本文结合大脑网络的连接特性,通过改变人工神经网络的连接实现网络的高效性.实验结果表明,结合生物脑网络的连接特性,改变网络的连接,很大程度上减少了网络的计算成本,而网络的性能并没有受到明显影响.     

基于TEEN协议和BP神经网络的WSN数据融合模型

针对无线传感网感知数据中含有大量无效或冗余数据的现象,本文提出了一种基于TEEN协议和BP(Back Propagation, BP)神经网络的数据融合模型。该模型利用三层BP神经网络描述簇结构,通过TEEN阈值过滤非必要信息,在簇结构信息传输过程中运用神经网络功能函数处理大量感知数据,从中提取感知数据的特征值并转发至汇聚节点。实验仿真表明,该模型无论在数据通信量、使用寿命及网络消耗上都优于TEEN协议,在降低网络通信量和网络能耗的同时提升了网络的使用寿命,大大提升了数据采集的效率和性能。     

Computer networking at the National Computing Centre

The National Computing Centre installed a microcomputer-based local area network in 1981. The paper reviews the operation of this network since its implementation. The configuration of the network and the facilities it provides are described. The justification for such a network is examined, and the objectives of the installation are listed. The criteria on which the network was chosen are discussed, and implementation, training and support are described. The applications of the network are examined, along with user reaction, the impact on work and the benefits achieved. Problems and limitations of the network are also discussed.     

网络流驱动的复杂输运网络演化模型

传统复杂网络演化模型在网络拓扑结构与边权演化的机制设计中,未考虑网络流对于输运网络演化的驱动作用。引入网络流的动态驱动机制,分析网络流的规模增长、空间距离的制约与最短路径的配流策略三种因素,在输运网络的演化过程中所发挥的作用。发现这三种因素并不足以改变复杂网络的无标度性;基于最短路径的配流机制是网络流分布不均的关键影响因素;空间距离抑制作用对于网络相配性具有关键影响作用。     

无线有组织网络

无线有组织网络是网络节点间有内在关系的无基础设施的无线网络,其除具有无基础设施无线网络的特征外,网络有中心节点,且节点位置关系相对固定。这种特殊的网络应用于野外救援、军事活动等临时性通信。首先回顾了对讲机系统、无线分组网络和无线自组织网络等无基础设施无线网络的结构和特点,它们都不能包涵无线有组织网络特征。详细定义了无线有组织网的相关模型,分析了无线有组织网络的特点和研究重点。     

复杂网络生存性评估方法研究

该文针对复杂网络的特点,首先给出了复杂网络生存性的一个新测度——容忍度。在此基础之上,给出了生存性测度的新定义,针对复杂网络无标度性的特点,给出了复杂网络生存性评估的新方法,并以互联网抽样数据为例进行了网络生存性分析。最后对复杂网络生存性研究的思路进行了探讨,指出从网络拓扑结构出发,研究拓扑结构的各种属性对网络生存性的影响,将是复杂网络生存性研究的一个有效而新颖的思路。     

计算一类有向网络可靠性的线性时间算法

该文使用的可靠性保护缩减的方法计算有向网络ST可靠性（存在从源点到汇点正常运行道路的概率）是计算网络可靠性的常用方法之一,而且人们非常关心怎样的网络计算其可靠性存在线性时间算法,作者提出了两类新的可靠性保护缩减－－源桥缩减和惠斯通桥缩减和一类有向无圈网络,称之为WST网络,该类网络是对以前的BSP网络的扩展并且对于该类网络提出了一个计算其可靠性的线性时间算法。     

基于Hopfield神经网络算法的WSN路径优化

针对无线传感器网络（WSN）能量有限的特点,提出一种新的基于Hopfield神经网络的路由优化算法,同时给出能量函数各参数之间的关系。通过Matlab软件对不同规模的网络进行仿真,仿真结果表明,该算法是可行的。     

基于多层网络的供应链金融建模与分析*

以多层网络理论为理论基础,将企业看成节点,企业间的关系看成连边,依据供应链金融真实的交易背景,建立供应链金融网络演化模型,试图为供应链金融提供新的研究方法。通过数学解析得到该供应链金融网络累积稳态平均度分布,利用计算机模拟对结果进行分析。结果表明多层网络建模得到的累积稳态平均度分布指数大于3,供应链金融网络呈无标度特性,供应链金融网络对供应链关系的依赖程度以及供应链金融网络新节点的连边数会改变网络的标度律大小,影响中枢节点的数量。     

Banet — a local area network for distributed data processing

A broadcast architecture network (Banet) suitable for distributed data processing is proposed. One feature of Banet is that the broadcast-within-a-group function is supported not only by the datalink level but by the transport or session-level network structure. The commitment control scheme is included in the network protocol. Design goals, physical structure and protocols are discussed.     

一种基于好友簇的社交网络中的时延容忍路由协议

社交网络是一种以便携式移动通信设备为节点的无线网络,通常由于其规模较大、结构复杂并且拓扑变化频繁,而成为时延容忍网络的一个典型应用场景。通过分析社交网络的特性,构建了基于好友群组的网络拓扑模型,并基于该模型,提出了一种基于簇结构的时延容忍路由协议。通过实验证明了该路由协议可以在保证较高路由性能的前提下有效控制由于数据副本传染造成的对网络资源的消耗。     

基于无线传感器网络的智能家居系统

无线传感器网络通过各类集成化的微型传感器协作地实时监测、感知和采集各种环境或监测对象的信息;而智能家居系统的设计是要在家庭内部以家庭总线的方式构建一个相当于住宅神经的家庭控制网络;本文把传感器网络引入到智能家居中,设计了一个基于传感器网络的智能家居系统。     

Wavelet networks for sensor signal classification in flank wear assessment

It is known that the force and vibration sensor signals in a turning process are sensitive to the gradually increasing flank wear. Based on this fact, this paper investigates a flank wear assessment technique in turning through force and vibration signals. Mainly to reduce the computational burden associated with the existing sensor-based methods for flank wear assessment, a so-called wavelet network is investigated. The basic idea in this new method is to optimize simultaneously the wavelet parameters (that represent signal features) and the signal-interpretation parameters (that are equivalent to neural network weights) to eliminate the feature extraction phase without increasing the computational complexity of the neural network. A neural network architecture similar to a standard one-hidden-layer feedforward neural network is used to relate sensor signal measurements to flank wear classes. A novel training algorithm for such a network is developed. The performance of this n ew method is compared with a previously developed flank wear assessment method which uses a separate feature extraction step. The proposed wavelet network can also be useful for developing signal interpretation schemes for manufacturing process monitoring, critical component monitoring, and product quality monitoring.     

A cognitive scheme for gateway protection in wireless sensor network

In wireless sensor networks, sensor readings are gathered at a gateway for processing and forwarding to a remote command center. The potential closeness of the gateway to dangerous events, e.g. fires, exposes it to damage and thus risks making the network dysfunctional. Therefore, protecting the gateway by repositioning it away from safety-hazardous spots is critical for the operation of the network. However, moving the gateway too far from the sensors that report on active events would have negative effect on the network performance, e.g. throughput and energy consumption. Therefore, balancing the gateway safety and network performance goals will be necessary. In this paper, we present GRISP, a novel Gateway Relocation algorithm for Improved Safety and Performance. GRISP employs an evolutionary neural network model to assess the safety of the gateway at the various locations. The model is then used to direct the search in an area of interest for a safer position that would enhance or at least maintain an acceptable level of network performance. In addition, GRISP guides the gateway during the move by finding safe paths leading to the new location. Our experimental validation results demonstrate the effectiveness of GRISP.     

Design and analysis of an efficient neural network model for solving nonlinear optimization problems

This paper presents an efficient approach based on a recurrent neural network for solving constrained nonlinear optimization. More specifically, a modified Hopfield network is developed, and its internal parameters are computed using the valid-subspace technique. These parameters guarantee the convergence of the network to the equilibrium points that represent an optimal feasible solution. The main advantage of the developed network is that it handles optimization and constraint terms in different stages with no interference from each other. Moreover, the proposed approach does not require specification for penalty and weighting parameters for its initialization. A study of the modified Hopfield model is also developed to analyse its stability and convergence. Simulation results are provided to demonstrate the performance of the proposed neural network.     

一种基于GAN的异构信息网络表示学习方法

异构信息网络中包含丰富的结构和语义信息,通过网络表示学习保留异构信息网络的结构和语义信息是当前研究的热点。传统的异构信息网络表示学习方法局限于利用元路径的形式保留异构信息网络中的语义信息,缺乏考虑网络中所有节点的分布情况,保留的信息不够充分。因此,本文提出一种基于生成式对抗网络（Generative Adversarial Networks, GAN）的异构信息网络表示学习方法（HINGAN）,其能更好地保留网络中的结构信息和语义信息。HINGAN中通过生成模型和判别模型的对抗学习,提高表示学习的鲁棒性。基于2个真实数据集的实验结果表明,本文提出的模型与传统的异构信息网络方法相比,在节点分类和链接预测任务中的结果都有明显提升。     

Currents in connectionism

This paper reviews four significant advances on the feedforward architecture that has dominated discussions of connectionism. The first involves introducing modularity into networks by employing procedures whereby different networks learn to perform different components of a task, and a Gating Network determines which network is best equiped to respond to a given input. The second consists in the use of recurrent inputs whereby information from a previous cycle of processing is made available on later cycles. The third development involves developing compressed representations of strings in which there is no longer an explicit encoding of the components but where information about the structure of the original string can be recovered and so is present functionally. The final advance entails using connectionist learning procedures not just to change weights in networks but to change the patterns used as inputs to the network. These advances significantly increase the usefulness of connectionist networks for modeling human cognitive performance by, among other things, providing tools for explaining the productivity and systematicity of some mental activities, and developing representations that are sensitive to the content they are to represent.A version of this paper was the invited address at the first annual conference of the Society for Machines and Mentality in December, 1991. A related version was presented at the Computers and Philosophy Conference in August, 1992. I am extremely grateful to Adele Abrahamsen for her numerous suggestions and comments on various drafts of this paper.     

Credal网络推理的一种不完全枚举法

Credal网络是研究不确定环境下知识表示和因果推理的一种图模型,其条件概率值可以用不精确的区间或不等式定性地表示,使得表达方式更加灵活有效。Credal网络的推理是计算一定证据下的后验概率最大值和最小值,给出了一种Credal网络推理的新方法,该方法是在桶消元框架下通过枚举计算部分因子函数值,使计算量大大减小,并且可以得到精确的结果。最后用一个实例说明了该方法的可行性。     

Two-node-Hamiltonicity of enhanced pyramid networks

An enhanced pyramid network is an alternate hierarchical structure for a pyramid network. This structure is created in a pyramid network by replacing each mesh with a torus at layers greater than one. This work studies the fault-tolerant Hamiltonian problem on the enhanced pyramid network and demonstrates that an enhanced pyramid network with two faulty nodes is Hamiltonian. The result is optimal, because edge connectivity and node connectivity of the enhanced pyramid network are both 4.     

A hub location problem with fully interconnected backbone and access networks

This paper considers the design of two-layered fully interconnected networks. A two-layered network consists of clusters of nodes, each defining an access network and a backbone network. We consider the integrated problem of determining the access networks and the backbone network simultaneously. A mathematical formulation is presented, but as the linear programming relaxation of the mathematical formulation is weak, a formulation based on the set partitioning model and column generation approach is also developed. The column generation subproblems are solved by solving a series of quadratic knapsack problems. We obtain superior bounds using the column generation approach than with the linear programming relaxation. The column generation method is therefore developed into an exact approach using the branch-and-price framework. With this approach we are able to solve problems consisting of up to 25 nodes in reasonable time. Given the difficulty of the problem, the results are encouraging.