Dynamical complexity in the C.elegans neural network

We model the neuronal circuit of the C.elegans soil worm in terms of a Hindmarsh-Rose system of ordinary differential equations, dividing its circuit into six communities which are determined via the Walktrap and Louvain methods. Using the numerical solution of these equations, we analyze important measures of dynamical complexity, namely synchronicity, the largest Lyapunov exponent, and the Φ_AR auto-regressive integrated information theory measure. We show that Φ_AR provides a useful measure of the information contained in the C.elegans brain dynamic network. Our analysis reveals that the C.elegans brain dynamic network generates more information than the sum of its constituent parts, and that attains higher levels of integrated information for couplings for which either all its communities are highly synchronized, or there is a mixed state of highly synchronized and desynchronized communities.     

Google matrix of Bitcoin network

We construct and study the Google matrix of Bitcoin transactions during the time period from the very beginning in 2009 till April 2013. The Bitcoin network has up to a few millions of bitcoin users and we present its main characteristics including the PageRank and CheiRank probability distributions, the spectrum of eigenvalues of Google matrix and related eigenvectors. We find that the spectrum has an unusual circle-type structure which we attribute to existing hidden communities of nodes linked between their members. We show that the Gini coefficient of the transactions for the whole period is close to unity showing that the main part of wealth of the network is captured by a small fraction of users. In global the Google matrix analysis of bitcoin network gives a new understanding of the bitcoin transactions with PageRank and CheiRank characterization of sellers and buyers which are dominant not simply due to the sold/bought volume but also by taking into account if bitcoins are sold to (bought by) other important sellers (buyers).     

Google matrix of business process management

Development of efficient business process models and determination of their characteristic properties are subject of intense interdisciplinary research. Here, we consider a business process model as a directed graph. Its nodes correspond to the units identified by the modeler and the link direction indicates the causal dependencies between units. It is of primary interest to obtain the stationary flow on such a directed graph, which corresponds to the steady-state of a firm during the business process. Following the ideas developed recently for the World Wide Web, we construct the Google matrix for our business process model and analyze its spectral properties. The importance of nodes is characterized by PageRank and recently proposed CheiRank and 2DRank, respectively. The results show that this two-dimensional ranking gives a significant information about the influence and communication properties of business model units. We argue that the Google matrix method, described here, provides a new efficient tool helping companies to make their decisions on how to evolve in the exceedingly dynamic global market.     

Image noise cancellation using linear matrix inequality and cellular neural network

In this paper, the technique of image noise cancellation is presented by employing cellular neural networks (CNN) and linear matrix inequality (LMI). The main objective is to obtain the templates of CNN by using a corrupted image and a corresponding desired image. A criterion for the uniqueness and global asymptotic stability of the equilibrium point of CNN is presented based on the Lyapunov stability theorem (i.e., the feedback template “A” of CNN is solved at this step), and the input template “B” of CNN is designed to achieve desirable output by using the property of saturation nonlinearity of CNN. It is shown that the problem of image noise cancellation can be characterized in terms of LMIs. The simulation results indicate that the proposed method is useful for practical application.     

Time partition function analysis of a neural network

The time characteristics of a linear network in the brain are obtained by the method of the time partition function, which is analogous to a grand partition function or a distribution function in statistical mechanics. The analogy between the average density in a many-particle system and the reciprocal of the frequency in a network is shown. By this method, the frequency distribution functions are obtained with respect to a network composed of two layers, the network used in information retrieval and the network generating a brain wave.     

Spectral properties of Google matrix of Wikipedia and other networks

We study the properties of eigenvalues and eigenvectors of the Google matrix of the Wikipedia articles hyperlink network and other real networks. With the help of the Arnoldi method, we analyze the distribution of eigenvalues in the complex plane and show that eigenstates with significant eigenvalue modulus are located on well defined network communities. We also show that the correlator between PageRank and CheiRank vectors distinguishes different organizations of information flow on BBC and Le Monde web sites.     

Image watermarking capacity analysis based on Hopfield neural network

In watermarking schemes, watermarking can be viewed as a form of communication problems. Almost all of previous works on image watermarking capacity are based on information theory, using Shannon formula to calculate the capacity of watermarking. In this paper, we present a blind watermarking algorithm using Hopfield neural network, and analyze watermarking capacity based on neural network. In our watermarking algorithm, watermarking capacity is decided by attraction basin of associative memory.     

Data analysis in relativistic nuclear physics using neural network

Two neural network algorithms for data analysis in relativistic nuclear physics are presented. A neural network technique (Hopfield method) is used in order to reconstruct particle tracks starting from a data set obtained with a coordinate detector system. An algorithm for circles recognition using deformable templates is carried out and its performances are studied. The technical limitations of the detectors, which in real situation prevent the possibility to reconstruct hits right on the circle, and presence of the noise points are taken into account.     

Koopman analysis of nonlinear systems with a neural network representation

The observation and study of nonlinear dynamical systems has been gaining popularity over years in different fields. The intrinsic complexity of their dynamics defies many existing tools based on individual orbits, while the Koopman operator governs evolution of functions defined in phase space and is thus focused on ensembles of orbits, which provides an alternative approach to investigate global features of system dynamics prescribed by spectral properties of the operator. However, it is difficult to identify and represent the most relevant eigenfunctions in practice. Here, combined with the Koopman analysis, a neural network is designed to achieve the reconstruction and evolution of complex dynamical systems. By invoking the error minimization, a fundamental set of Koopman eigenfunctions are derived, which may reproduce the input dynamics through a nonlinear transformation provided by the neural network. The corresponding eigenvalues are also directly extracted by the specific evolutionary structure built in.     

卷积神经网络的缺陷类型识别分析

该文提出一种基于卷积神经网络直接对阵列超声检测原始信号进行缺陷类型识别的方法,该方法无需对超声回波原始信号进行特征提取.文章研究对比了不同卷积神经网络及其优化的识别性能.首先采用超声相控阵系统对不同试块上的平底孔、球底孔、通孔三种缺陷进行超声检测,然后利用LeNet5、VGG16和ResNet三种卷积神经网络对一维和二...     

Quantum neural network

In the neural network theory content-addressable memories are defined by patterns that are attractors of the dynamical rule of the system. This paper develops a quantum neural network starting from a classical neural network Hamiltonian and using a Schrödinger-like equation. It then shows that such a system exhibits probabilistic memory storage characteristics analogous to those of the dynamical attractors of classical systems.     

利用几何特性及神经网络进行人脸探测技术的研究

在人脸识别过程中 ,首先也是最重要的一个环节是人脸探测 ,因为一旦从图像中定位并提取到了人脸 ,那么下一步的人脸识别工作就变得非常容易。眼睛是人脸图像中最容易探测的部位 ,而且通过探测双眼来发现人脸最符合人的视觉习惯。提出了一种基于几何特征分析和人工神经网络的由粗到细的两级人脸探测方法。在第一级中 ,眼睛和脸是通过测量眼睛的尺寸和眼睛与脸的位置关系探测到的 ,第一级的输出是一个尺寸归一化的人脸 ,但偶尔也伴随着一个或多个因对复杂背景中与眼睛类似的物体的误判而得到的非人脸图像 ;第二级神经网络正是用来过滤掉第一级中被误判的人脸。实验表明 ,这种由粗到细的两级人脸探测系统具有很高的稳定性和探测正确率     

Acoustic attenuation analysis of network systems by using impedance matrix method

An impedance matrix method is proposed to predict the acoustic attenuation characteristics of network systems. The system may contain several silencer modules and each module may be composed of complex components such as multiply connected tubes, portions with any-shaped cross-section and dissipative parts. The technique of substructuring is adopted and the system is divided into several substructure modules. Three strategies: boundary element method (BEM), numerical point collocation approach and numerical mode matching technique are introduced and the impedance matrix of each module may be computed by a certain appropriate methodology according to the dimensions and geometry of the substructure. Impedance matrix synthesis is employed to obtain the resultant impedance matrix and then transmission loss may be calculated. All the calculation results are verified by experimental measurements and 3-D BEM predictions.     

MAP kinase cascades regulating axon regeneration in C. elegans

Mitogen-activated protein kinase (MAPK) signaling cascades are activated by diverse stimuli such as growth factors, cytokines, neurotransmitters and various types of cellular stress. Our evolving understanding of these signal cascades has been facilitated by genetic analyses and physiological characterization in model organisms such as the nematode Caenorhabditis elegans. Genetic and biochemical studies in C. elegans have shed light on the physiological roles of MAPK cascades in the control of cell fate decision, neuronal function and immunity. Recently it was demonstrated that MAPK signaling is also important for axon regeneration in C. elegans, and the use of C. elegans as a model system has significantly advanced our understanding of the largely conserved molecular mechanisms underlying axon regeneration. This review summarizes our current understanding of the role and regulation of MAPK signaling in C. elegans axon regeneration.     

基于脉冲耦合神经网络的拉曼光谱定性分析

通过对脉冲耦合神经网络(pulse coupled neural network,PCNN)和拉曼光谱定性分析的研究,提出了基于PCNN的拉曼光谱定性分析方法.首先,利用PCNN神经元的疲劳与不应期特性将拉曼光谱数据进行编码;然后,基于改进的Horspool算法将检测样品对应编码与基码数据库中的所有基码逐一匹配,并得到各对应的匹配相似度,进而判定样品类别.相关实验和数据分析证明了该文方法的准确性和有效性.同时,该文方法避免了目前基于谱模版定性分析方法中待测样品拉曼光谱特征谱峰难以确定以及匹配分析冗余度高等不足,且对存储空间的要求仅为后者的5.8%.     

An linear matrix inequality approach to global synchronisation of non-parameter perturbations of multi-delay Hopfield neural network

In this study,a successful linear matrix inequality approach is used to analyse a non-parameter perturbation of multi-delay Hopfield neural network by constructing an appropriate Lyapunov-Krasovskii functional.This paper presents the comprehensive discussion of the approach and also extensive applications.     

Analysis of world terror networks from the reduced Google matrix of Wikipedia

We apply the reduced Google matrix method to analyze interactions between 95 terrorist groups and determine their relationships and influence on 64 world countries. This is done on the basis of the Google matrix of the English Wikipedia (2017) composed of 5 416 537 articles which accumulate a great part of global human knowledge. The reduced Google matrix takes into account the direct and hidden links between a selection of 159 nodes (articles) appearing due to all paths of a random surfer moving over the whole network. As a result we obtain the network structure of terrorist groups and their relations with selected countries including hidden indirect links. Using the sensitivity of PageRank to a weight variation of specific links we determine the geopolitical sensitivity and influence of specific terrorist groups on world countries. The world maps of the sensitivity of various countries to influence of specific terrorist groups are obtained. We argue that this approach can find useful application for more extensive and detailed data bases analysis.     

Pattern-specific neural network design

We present evidence that the performance of the traditional fully connected Hopfield model can be dramatically improved by carefully selecting an information-specific connectivity structure, while the synaptic weights of the selected connections are the same as in the Hopfield model. Starting from a completely disconnected network we let genuine Hebbian synaptic connections grow, one by one, until a desired degree of stability is achieved. Neural pathways are thus fixed notbefore, butduring the learning phase.     

Global Hopf bifurcation analysis on a BAM neural network with delays

A delayed differential equation that models a bidirectional associative memory (BAM) neural network with four neurons is considered. By using a global Hopf bifurcation theorem for FDE and a Bendixon's criterion for high-dimensional ODE, a group of sufficient conditions for the system to have multiple periodic solutions are obtained when the sum of delays is sufficiently large.     

Vibrational analysis of carbon nanotubes using molecular mechanics and artificial neural network

This paper presents the molecular mechanics based finite element modeling of carbon nanotubes (CNTs) and their applications as mass sensors. The beam element with elastic behavior is considered as the bond between the carbon atoms and its properties are obtained using equating continuum and molecular characteristics. The first five natural frequencies of CNTs in cantilever and doubly clamped boundary conditions (BCs) and their corresponding mode shapes are studied in detail. Furthermore, a multilayer perceptron neural network is used to predict the fundamental vibration frequencies of the CNTs with different diameters and lengths. In addition, variations of the natural frequencies of the CNTs with distorted cross sections are investigated. Moreover, the effects of some attached masses with various values on the first three natural frequencies of a considered CNT are studied here.