Nonparametric statistics of dynamic networks with distinguishable nodes

The study of random graphs and networks had an explosive development in the last couple of decades. Meanwhile, techniques for the statistical analysis of sequences of networks were less developed. In this paper, we focus on networks sequences with a fixed number of labeled nodes and study some statistical problems in a nonparametric framework. We introduce natural notions of center and a depth function for networks that evolve in time. We develop several statistical techniques including testing, supervised and unsupervised classification, and some notions of principal component sets in the space of networks. Some examples and asymptotic results are given, as well as two real data examples.     

Shear thickening of F-actin networks crosslinked with non-muscle myosin IIB

The material properties of cytoskeletal F-actin networks facilitate a broad range of cellular behaviors, whereby in some situations cell shape is preserved in the presence of force and, at other times, force results in irreversible shape change. These behaviors strongly suggest that F-actin networks can variably deform elastically or viscously. While a significant amount is known about the regulation of the elastic stiffness of F-actin networks, our understanding of the regulation of viscous behaviors of F-actin networks is largely lacking. Here, we study the rheological behavior of F-actin networks formed with heavy meromyosin non-muscle IIB (NMMIIB). We show that NMMIIB quenched with ADP crosslinks F-actin into networks that, for sufficient densities, display stress stiffening behavior. By performing a series of creep tests, we show that densely crosslinked actin/NMMIIB-ADP networks undergo viscous deformation over a wide range of stresses, ranging from 0.001 to 10 Pa. At high stresses, networks that stress stiffen are also observed to shear thicken, whereby the effective viscosity increases as a function of stress. Shear thickening results in a reduction in the extent of irreversible, viscous deformation in actin/NMMIIB-ADP networks at high stresses compared to that expected for a linear viscoelastic material. Thus, viscous deformation contributes less to the overall mechanical response at high levels of applied force. Our results indicate mechanisms by which the fluid-like nature of the actomyosin cytoskeleton can be reduced under high load.     

永磁同步电机系统、网络混沌振荡同步中断分析与控制

已有研究表明,电机在某些参数及工作条件下会出现混沌振荡行为。另一方面,现代工业生产过程中,经常需要多台电机协调同步运行。当前,国内外对电机系统、网络的混沌振荡同步分析与控制已有较多研究,但对电机系统、网络由于参数扰动引起同步中断的分析与控制研究尚属少见。为此,在传统的扩散耦合方式基础上,通过引入单向交叉耦合方式,研究了永磁同步电机(PMSM)系统、网络受到参数扰动后的振动特性。分析并证明了单向交叉耦合对PMSM系统、网络的参数扰动具有同步鲁棒性;并以3节点、4节点和一个多节点网络的PMSM系统验证,交叉耦合对多节点电机网络全局稳定同步的重要作用。研究结果表明,通过建立合适的单向交叉耦合方式,能够有效控制电机网络因参数扰动而引发的同步中断,使同步中断得以重新恢复。     

Simulation studies of multiclass queueing networks

Multiclass queueing networks have been used to model communication and complex manufacturing systems. Recent results have shown that these networks can be unstable in the sense that the total number of jobs in the network explodes as time tends to infinity even if the traffic intensity at each station is less than unity. Simulation is used to demonstrate these unusual phenomena. Conjectures based on simulation are presented.     

Interaction intimacy organizes networks of antagonistic interactions in different ways

Interaction intimacy, the degree of biological integration between interacting individuals, shapes the ecology and evolution of species interactions. A major question in ecology is whether interaction intimacy also shapes the way interactions are organized within communities. We combined analyses of network structure and food web models to test the role of interaction intimacy in determining patterns of antagonistic interactions, such as host–parasite, predator–prey and plant–herbivore interactions. Networks describing interactions with low intimacy were more connected, more nested and less modular than high-intimacy networks. Moreover, the performance of the models differed across networks with different levels of intimacy. All models reproduced well low-intimacy networks, whereas the more elaborate models were also capable of reproducing networks depicting interactions with higher levels of intimacy. Our results indicate the key role of interaction intimacy in organizing antagonisms, suggesting that greater interaction intimacy might be associated with greater complexity in the assembly rules shaping ecological networks.     

一类具比例时滞细胞神经网络反周期解的指数稳定性

本文针对一类具比例时滞细胞神经网络反周期解的全局指数稳定性进行研究.首先利用非线性变换将一类具比例时滞的细胞神经网络等价变换成一类具常时滞变系数的细胞神经网络.然后通过构造合适的时滞微分不等式和利用不等式技巧,得到了保证该系统反周期解的存在性和全局指数稳定性时滞依赖的充分条件.最后数值算例结果验证所得结论的正确性和与以往文献相比较低的保守性.     

Citation choice and innovation in science studies

What are the factors which render an article more likely to be cited? Using social network analysis of citations between published scholarly works, the nascent field around Social Studies of Science is examined from its incipience in 1971 until 2008. To gauge intellectual positioning, closeness centrality and orthodoxy rates are derived from bibliographic networks. Bibliographic orthodoxy is defined as the propensity of an article to cite other highly popular works. Orthodoxy and closeness centrality have differing effects on citation rates, varying across historical periods of development in the field. Effects were modest, but significant. In early time periods, articles with higher orthodoxy rates were cited more, but this effect dissipated over time. In contrast, citations associated with closeness centrality increased over time. Early SSS citation networks were smaller, less structurally cohesive and less modular than later networks. In contrast, later networks were larger, more structurally cohesive, more modular and less dense. These changes to the global SSS knowledge networks are linked to changes in the scientific reward structure ensconced in the network, particularly regarding orthodoxy and closeness centrality.     

基于人工神经网络的无芯片电子标签识别

本文提出了一种基于人工神经网络的无芯片电子标签识别方法.首先利用CST对不同夹角的标签进行仿真,得到水平和垂直两个方向的散射场,然后建立识别系统的神经网络模型,并利用该模型实现对标签的识别.该方法的优点是,使用1/2的样本数据训练神经网络模型,就可以得到准确的识别结果.仿真结果表明,基于神经网络的识别方法的误差在5°的范围内.     

Topological properties of robust biological and computational networks

Network robustness is an important principle in biology and engineering. Previous studies of global networks have identified both redundancy and sparseness as topological properties used by robust networks. By focusing on molecular subnetworks, or modules, we show that module topology is tightly linked to the level of environmental variability (noise) the module expects to encounter. Modules internal to the cell that are less exposed to environmental noise are more connected and less robust than external modules. A similar design principle is used by several other biological networks. We propose a simple change to the evolutionary gene duplication model which gives rise to the rich range of module topologies observed within real networks. We apply these observations to evaluate and design communication networks that are specifically optimized for noisy or malicious environments. Combined, joint analysis of biological and computational networks leads to novel algorithms and insights benefiting both fields.     

A privacy-preserving authentication protocol with secure handovers for the LTE/LTE-A networks

Long-Term Evolution/Long-Term Evolution Advanced (LTE/LTE-A) is the latest mobile communication technology that is offering high data rates and robust performance to the subscribers. Since LTE/LTE-A standards are established on the Internet Protocol (IP) connectivity and provide compatibility with the heterogeneous networks, these new features create availability of the new security challenges in the LTE/LTE-A networks. Taking into consideration the issues of serious signalling congestion and security loopholes in LTE/LTE-A networks, the authors propose an Efficient Authentication and Key Agreement Protocol for Evolved Packet System (EAKA-EPS) with secure handover procedures. The proposed protocol achieves outstanding results in terms of the optimization of computation and signalling overhead. With this, the protocol guarantees the needed security requirements like protected wireless interface and strong mutual authentication between the entities, and ensures access stratum secrecy at the time of handovers. The formal verification results of the proposed scheme over the security verification and simulation tool “Automated Validation of Internet Security Protocols and Applications (AVISPA)” show that the suggested protocol is safe against various malicious attacks, which are still possible in LTE/LTE-A networks. To the best of the authors’ knowledge, the suggested approach is the first approach that provides perfect secrecy with less computation and communication overhead in the LTE/LTE-A networks.     

Fatigue life prediction of composite materials using polynomial classifiers and recurrent neural networks

Due to their massively parallel structure and ability to learn by example, artificial neural networks can deal with nonlinear problems for which an accurate analytical solution is difficult to obtain. These networks have been used in modeling the mechanical behavior of fiber-reinforced composite materials. Although promising results were obtained using such networks, more investigation on the appropriate choice of their structure and their performance in the presence of limited and noisy data is needed. On the other hand, polynomials networks have been known to have excellent properties as classifiers and are universal approximators to the optimal Bayes classifier. Not being dependant on various user defined parameters, having less computational requirements makes their use over other methods, such as neural networks, an advantage.

In this work, the fatigue behavior of unidirectional glass fiber/epoxy composite laminae under tension–tension and tension–compression loading is predicted using feedforward and recurrent neural networks. These predictions are compared to those obtained using polynomial classifiers. Experimental data obtained for fiber orientation angles of 0°, 19°, 45°, 71° and 90° under stress ratios of 0.5, 0 and –1 is used.

It is shown that, even when a small number of experimental data points is used to train both polynomial classifiers and neural networks, the predictions obtained are comparable to other current fatigue life-prediction methods. Also, polynomial classifiers are shown to provide accurate modeling between the input parameters (maximum stress, R-ratio, fiber orientation angle) and the number of cycles to failure when compared to neural networks.     

M‐matrix‐based stability conditions for genetic regulatory networks with time‐varying delays and noise perturbations

Stability is essential for designing and controlling any dynamic systems. Recently, the stability of genetic regulatory networks has been widely studied by employing linear matrix inequality (LMI) approach, which results in checking the existence of feasible solutions to high‐dimensional LMIs. In the previous study, the authors present several stability conditions for genetic regulatory networks with time‐varying delays, based on M ‐matrix theory and using the non‐smooth Lyapunov function, which results in determining whether a low‐dimensional matrix is a non‐singular M ‐matrix. However, the previous approach cannot be applied to analyse the stability of genetic regulatory networks with noise perturbations. Here, the authors design a smooth Lyapunov function quadratic in state variables and employ M ‐matrix theory to derive new stability conditions for genetic regulatory networks with time‐varying delays. Theoretically, these conditions are less conservative than existing ones in some genetic regulatory networks. Then the results are extended to genetic regulatory networks with time‐varying delays and noise perturbations. For genetic regulatory networks with n genes and n proteins, the derived conditions are to check if an n × n matrix is a non‐singular M ‐matrix. To further present the new theories proposed in this study, three example regulatory networks are analysed.Inspec keywords: genetics, linear matrix inequalities, Lyapunov matrix equations, molecular biophysics, noise, proteinsOther keywords: M‐matrix‐based stability condition, genetic regulatory networks, time‐varying delays, noise perturbations, linear matrix inequality approach, high‐dimensional LMI, Lyapunov function, state variables, M‐matrix theory, proteins, nonsingular M‐matrix     

The resilience of regional African HIV/AIDS research networks to the withdrawal of international authors in the subfield of public administration and governance: lessons for funders and collaborators

This paper compares the position and performance of Africans in international research networks, comparing a well-funded and internationally driven research network against a network developing organically with less funding and oversight. Specifically, we map the co-authorship networks related to African governance and public policy (1) in general and (2) related to HIV/AIDS. In both research networks Africans are well positioned globally, but African authors have lower connectivity and status in the HIV/AIDS network than the general network. Links between authors in different African countries are often bridged by non-Africans. This makes the African research networks vulnerable to shifting funding priorities; and international dropout would fragment the HIV/AIDS network more than the general network. We conclude that funders and researchers should prioritize direct inter-country African collaborations to improve the resilience of African research networks.     

Identification of Factors Governing Mechanical Properties of TRIP-Aided Steel Using Genetic Algorithms and Neural Networks

Mechanical properties of transformation induced plasticity (TRIP)-aided multiphase steels are modeled by neural networks using two methods of reducing the network connectivity, viz. a pruning algorithm and a predator prey algorithm, to gain understanding on the impact of steel composition and treatment. The pruning algorithm gradually reduces the complexity of the lower layer of connections, removing less significant connections. In the predator prey algorithm, a genetic algorithm based multi-objective optimization technique evolves neural networks on a Pareto front, simultaneously minimizing training error and network size. The results show that the techniques find parsimonious models and, furthermore, extract useful knowledge from the data.     

FPGA Implementation of Deep Leaning Model for Video Analytics

In recent years, deep neural networks have become a fascinating and influential research subject, and they play a critical role in video processing and analytics. Since, video analytics are predominantly hardware centric, exploration of implementing the deep neural networks in the hardware needs its brighter light of research. However, the computational complexity and resource constraints of deep neural networks are increasing exponentially by time. Convolutional neural networks are one of the most popular deep learning architecture especially for image classification and video analytics. But these algorithms need an efficient implement strategy for incorporating more real time computations in terms of handling the videos in the hardware. Field programmable Gate arrays (FPGA) is thought to be more advantageous in implementing the convolutional neural networks when compared to Graphics Processing Unit (GPU) in terms of energy efficient and low computational complexity. But still, an intelligent architecture is required for implementing the CNN in FPGA for processing the videos. This paper introduces a modern high-performance, energy-efficient Bat Pruned Ensembled Convolutional networks (BPEC-CNN) for processing the video in the hardware. The system integrates the Bat Evolutionary Pruned layers for CNN and implements the new shared Distributed Filtering Structures (DFS) for handing the filter layers in CNN with pipelined data-path in FPGA. In addition, the proposed system adopts the hardware-software co-design methodology for an energy efficiency and less computational complexity. The extensive experimentations are carried out using CASIA video datasets with ARTIX-7 FPGA boards (number) and various algorithms centric parameters such as accuracy, sensitivity, specificity and architecture centric parameters such as the power, area and throughput are analyzed. These results are then compared with the existing pruned CNN architectures such as CNN-Prunner in which the proposed architecture has been shown 25% better performance than the existing architectures.     

Mapping the input–output relationship in HSLA steels through expert neural network

Modification of the architecture of the artificial neural network is done to accommodate the information available from the knowledge base in the field of materials science for thermomechanically processed HSLA steel. The complicated architectures of these networks are made to satisfy the well-understood physical metallurgy principles, which administer the property response to the combined actions of the compositional and process parameters. The networks developed have been found to give very good convergence during training. The number of epochs required to reach the targeted error was found less for these networks than the conventional networks.     

Cross-linked bacterial cellulose networks using glyoxalization

In this study, we demonstrate that bacterial cellulose (BC) networks can be cross-linked via glyoxalization. The fracture surfaces of samples show that, in the dry state, less delamination occurs for glyoxalized BC networks compared to unmodified BC networks, suggesting that covalent bond coupling between BC layers occurs during the glyoxalization process. Young's moduli of dry unmodified BC networks do not change significantly after glyoxalization. The stress and strain at failure are, however, reduced after glyoxalization. However, the wet mechanical properties of the BC networks are improved by glyoxalization. Raman spectroscopy is used to demonstrate that the stress-transfer efficiency of deformed dry and wet glyoxalized BC networks is significantly increased compared to unmodified material. This enhanced stress-transfer within the networks is shown to be a consequence of the covalent coupling induced during glyoxalization and offers a facile route for enhancing the mechanical properties of BC networks for a variety of applications.     

Source selection routing algorithms in integrated cellular networks

Integrated cellular networks (ICNs) are normally constructed by adding ad hoc overlay on cellular networks to solve the latter's flexibility and capacity expansion problems. In such networks, routing plays a critical role in finding a route to divert congested traffic from a congested cell to another less crowded cell. Much work has been conducted on routing protocols in ICNs, whereas no dedicated work has been found for an important aspect of routing, namely source selection. The process of a source selection can be an algorithm that is designed for selecting a proper pseudo- source to release its occupying channel to a blocked mobile user. Consequently, this pseudo-source diverts its ongoing call to another cell by using a free channel in a neighbour cell via a relaying route. Based on an introduction of a representative ICN infrastructure, three source selection algorithms are proposed. Both numerical analysis and evaluation results are presented, which show the efficiency of the algorithms and their different abilities in adapting to different network situations, such as traffic density and cell capacity.     

称重传感器蠕变误差的神经网络补偿方法

针对电阻应变式称重传感器存在严重的蠕变误差直接影响称重结果准确度的问题,提出了一种基于神经网络的称重传感器蠕变误差自动补偿模型,并给出了模型的训练算法。对量程为50 kg,C3等级的电阻应变式称重传感器进行了实验,实验结果表明,使用神经网络蠕变误差补偿方法补偿后,称重传感器加载标准砝码30 min内的蠕变误差最大变化量为0.0108 kg,小于国家标准GB/T 7551-200《称重传感器》规定的允许值。