Structure and Connectivity Analysis of Financial Complex System Based on G-Causality Network

The recent financial crisis highlights the inherent weaknesses of the financial market. To explore the mechanism that maintains the financial market as a system, we study the interactions of U.S. financial market from the network perspective. Applied with conditional Granger causality network analysis, network density, in-degree and out-degree rankings are important indicators to analyze the conditional causal relationships among financial agents, and further to assess the stability of U.S. financial systems. It is found that the topological structure of G-causality network in U.S. financial market changed in different stages over the last decade, especially during the recent global financial crisis. Network density of the G-causality model is much higher during the period of 2007-2009 crisis stage, and it reaches the peak value in 2008, the most turbulent time in the crisis. Ranked by in-degrees and out-degrees, insurance companies are listed in the top of 68 financial institutions during the crisis. They act as the hubs which are more easily influenced by other financial institutions and simultaneously influence others during the global financial disturbance.     

Granger Causality among Graphs and Application to Functional Brain Connectivity in Autism Spectrum Disorder

Graphs/networks have become a powerful analytical approach for data modeling. Besides, with the advances in sensor technology, dynamic time-evolving data have become more common. In this context, one point of interest is a better understanding of the information flow within and between networks. Thus, we aim to infer Granger causality (G-causality) between networks’ time series. In this case, the straightforward application of the well-established vector autoregressive model is not feasible. Consequently, we require a theoretical framework for modeling time-varying graphs. One possibility would be to consider a mathematical graph model with time-varying parameters (assumed to be random variables) that generates the network. Suppose we identify G-causality between the graph models’ parameters. In that case, we could use it to define a G-causality between graphs. Here, we show that even if the model is unknown, the spectral radius is a reasonable estimate of some random graph model parameters. We illustrate our proposal’s application to study the relationship between brain hemispheres of controls and children diagnosed with Autism Spectrum Disorder (ASD). We show that the G-causality intensity from the brain’s right to the left hemisphere is different between ASD and controls.     

Co-evolution of the brand effect and competitiveness in evolving networks

The principle that ‘the brand effect is attractive’ underlies the preferential attachment. Here we show that the brand effect is just one dimension of attractiveness. Another dimension is competitiveness. We firstly introduce a general framework that allows us to investigate the competitive aspect of real networks, instead of simply preferring popular nodes. Our model accurately describes the evolution of social and technological networks. The phenomenon that more competitive nodes become richer can help us to understand the evolution of many competitive systems in nature and society. In general,the paper provides an explicit analytical expression of degree distributions of the network. In particular, the model yields a nontrivial time evolution of nodes’ properties and the scale-free behavior with exponents depending on the microscopic parameters characterizing the competition rules. Secondly, through theoretical analyses and numerical simulations, we reveal that our model has not only the universality for the homogeneous weighted network, but also the character for the heterogeneous weighted network. Thirdly, we also develop a model based on the profit-driven mechanism. It can better describe the observed phenomenon in enterprise cooperation networks. We show that the standard preferential attachment,the growing random graph, the initial attractiveness model, the fitness model, and weighted networks can all be seen as degenerate cases of our model.     

Investigating directed cortical interactions in time-resolved fMRI data using vector autoregressive modeling and Granger causality mapping

We present a framework aimed to reveal directed interactions of activated brain areas using time-resolved fMRI and vector autoregressive (VAR) modeling in the context of Granger causality. After describing the underlying mathematical concepts, we present simulations helping to characterize the conditions under which VAR modeling and Granger causality can reveal directed interactions from fluctuations in BOLD-like signal time courses. We apply the proposed approach to a dynamic sensorimotor mapping paradigm. In an event-related fMRI experiment, subjects performed a visuomotor mapping task for which the mapping of two stimuli (“faces” vs “houses”) to two responses (“left” or “right”) alternated periodically between the two possible mappings. Besides expected activity in sensory and motor areas, a fronto-parietal network was found to be active during presentation of a cue indicating a change in the stimulus-response (S-R) mapping. The observed network includes the superior parietal lobule and premotor areas. These areas might be involved in setting up and maintaining stimulus-response associations. The Granger causality analysis revealed a directed influence exerted by the left lateral prefrontal cortex and premotor areas on the left posterior parietal cortex.     

Optimization of robustness of network controllability against malicious attacks

As the controllability of complex networks has attracted much attention recently, how to design and optimize the robustness of network controllability has become a common and urgent problem in the engineering field. In this work, we propose a method that modifies any given network with strict structural perturbation to effectively enhance its robustness against malicious attacks, called dynamic optimization of controllability. Unlike other structural perturbations, the strict perturbation only swaps the links and keeps the in- and out-degree unchanged. A series of extensive experiments show that the robustness of controllability and connectivity can be improved dramatically. Furthermore, the effectiveness of our method is explained from the views of underlying structure. The analysis results indicate that the optimization algorithm makes networks more homogenous and assortative.     

Characteristics of photonic bands generated by quadrangular multiconnected networks

In this paper, by means of the network equation and generalized dimensionless Floquet-Bloch theorem, we study the influences of the number of connected waveguide segments （NCWS） between adjacent nodes and the matching ratio of waveguide length （MRWL） on the photonic bands generated by quadrangular multiconnected networks （QMNs）, and obtain a series of formulae. It is found that multicombining networks （MCNs） and repetitive combining networks （RCNs） are equivalent to each other and they can all be simplified into the simplest fundamental combining systems. It would be useful for adjusting the number, widths, and positions of photonic bands, and would possess potential applications for the designing of all-optical devices and photonic network devices.     

Evolution Properties of Modules in Complex Networks

In complex networks, network modules play a center role, which carry out a key function. In this paper, we introduce the spatial correIation function to describe the relationships among the network modules. Our focus is to investigate how the network modules evolve, and what the evolution properties of the modules are. In order to test the proposed method, as the examples, we use our method to analyze and discuss the ER random network and scale-free network. Rigorous analysis of the existing data shows that the introduced correlation function is suitable for describing the evolution properties of network modules. Remarkably, the numerical simulations indicate that the ER random network and scale-free network have different evolution properties.     

Synchronization transition of a coupled system composed of neurons with coexisting behaviors near a Hopf bifurcation

The coexistence of a resting condition and period-1 firing near a subcritical Hopf bifurcation point, lying between the monostable resting condition and period-1 firing, is often observed in neurons of the central nervous systems. Near such a bifurcation point in the Morris-Lecar （ML） model, the attraction domain of the resting condition decreases while that of the coexisting period-1 firing increases as the bifurcation parameter value increases. With the increase of the coupling strength, and parameter and initial value dependent synchronization transition processes from non-synchronization to compete synchronization are simulated in two coupled ML neurons with coexisting behaviors： one neuron chosen as the resting condition and the other the coexisting period-1 firing. The complete synchronization is either a resting condition or period-1 firing dependent on the initial values of period-1 firing when the bifurcation parameter value is small or middle and is period- 1 firing when the parameter value is large. As the bifurcation parameter value increases, the probability of the initial values of a period- 1 firing neuron that lead to complete synchronization of period- 1 firing increases, while that leading to complete synchronization of the resting condition decreases. It shows that the attraction domain of a coexisting behavior is larger, the probability of initial values leading to complete synchronization of this behavior is higher. The bifurcations of the coupled system are investigated and discussed. The results reveal the complex dynamics of synchronization behaviors of the coupled system composed of neurons with the coexisting resting condition and period-1 firing, and are helpful to further identify the dynamics of the spatiotemporal behaviors of the central nervous system.     

Modeling Topology and Nonlinear Dynamical Behavior of the Weighted Scale-Free Networks

An improved weighted scale-free network, which has two evolution mechanisms： topological growth and strength dynamics, has been introduced. The topology structure of the model will be explored in details in this work. The evolution driven mechanism of Olami-Feder Christensen （OFC） model is added to our model to study the self-organlzed criticality and the dynamical behavior. We also.consider attack mechanism and the study of the model with attack is also investigated in this paper. We tlnd there are differences between the model with attack and without attack.     

Control of random sensitivity Boolean networks via of Boolean functions

In this paper, we discuss how to transform the disordered phase into an ordered phase in random Boolean networks. To increase the effectiveness, a control scheme is proposed, which periodically freezes a fraction of the network based on the average sensitivity of Boolean functions of the nodes. Theoretical analysis is carried out to estimate the expected critical value of the fraction, and shows that the critical value is reduced using this scheme compared to that of randomly freezing a fraction of the nodes. Finally, the simulation is given for illustrating the effectiveness of the proposed method.     

Weighted Evolving Networks with Self-organized Communities

In order to describe the self-organization of communities in the evolution of weighted networks, we propose a new evolving model for weighted community-structured networks with the preferential mechanisms functioned in different levels according to community sizes and node strengths, respectively. Theoretical analyses and numerical simulations show that our model captures power-law distributions of community sizes, node strengths, and link weights, with tunable exponents of v ≥ 1, γ 〉 2, and α 〉 2, respectively, sharing large clustering coefficients and scaling clustering spectra, and covering the range from disassortative networks to assortative networks. Finally, we apply our new model to the scientific co-authorship networks with both their weighted and unweighted datasets to verify its effectiveness.     

Self-organized Critical Model Based on Complex Brain Networks with Hierarchical Organization

The dynamical behavior in the cortical brain network of macaque is studied by modelling each cortical area with a subnetwork of interacting excitable neurons. We find that the avalanche of our model on different levels exhibits power-law. Furthermore the power-law exponent of the distribution and the average avalanche Size are affected by the topology of the network.     

Spiral Wave in Small-World Networks of Hodgkin-Huxley Neurons

The effect of small-world connection and noise on of Hodgkin-Huxley neurons are investigated in detail. Some the formation and transition of spiral wave in the networks interesting results are found in our numerical studies, i） The quiescent neurons are activated to propagate electric signal to others by generating and developing spiral wave from spiral seed in small area. ii） A statistical factor is defined to describe the collective properties and phase transition induced by the topology of networks and noise, iii） Stable rotating spiral wave can be generated and keeps robust when the rewiring probability is below certain threshold, otherwise, spiral wave can not be developed from the spiral seed and spiral wave breakup occurs for a stable rotating spiral wave. iv） Gaussian white noise is introduced on the membrane of neurons to study the noise-induced phase transition on spiral wave in small-world networks of neurons. It is confirmed that Ganssian white noise plays active role in supporting and developing spiral wave in the networks of neurons, and appearance of smaller factor of synchronization indicates high possibility to induce spiral wave.     

Effects of Vertex Activity and Self-organized Criticality Behavior on a Weighted Evolving Network

Effects of vertex activity have been analyzed on a weighted evolving network. The network is characterized by the probability distribution of vertex strength, each edge weight and evolution of the strength of vertices with different vertex activities. The model exhibits self-organized criticality behavior. The probability distribution of avalanche size for different network sizes is also shown. In addition, there is a power law relation between the size and the duration of an avalanche and the average of avalanche size has been studied for different vertex activities.     

An artificial neural network for proton identification in HERMES data

The HERMES time-of-flight （TOF） system is used for proton identification, but must be carefully calibrated for systematic biases in the equipment. This paper presents an artificial neural network （ANN） trained to recognize protons from ∧^0 decay using only raw event data such as time delay, momentum, and trajectory. To avoid the systematic errors associated with Monte Carlo models, we collect a sample of raw experimental data from the year 2000. We presume that when for a positive hadron （assigned one proton mass） and a negative hadron （assigned one π^- mass） the reconstructed invariant mass lies within the ∧^0 resonance, the positive hadron is more likely to be a proton. Such events are assigned an output value of one during the training process; all others were assigned the output value zero.
The trained ANN is capable of identifying protons in independent experimental data, with an efficiency equivalent to the traditional TOF calibration. By modifying the threshold for proton identification, a researcher can trade off between selection efficiency and background rejection power. This simple and convenient method is applicable to similar detection problems in other experiments.     

Global Synchronization of Directed Networks with Fast Switching Topologies

Global synchronization of a class of directed dynamical networks with switching topologies is investigated. It is found that if there is a directed spanning tree in the fixed time-average of network topology and the time-average is achieved sufficiently fast, then the network will reach global synchronization for sufficiently large coupling strength.     

Different Epidemic Models on Complex Networks

Models for diseases spreading are not just limited to SIS or SIR. For instance, for the spreading of AIDS/HIV, the susceptible individuals can be classified into different cases according to their immunity, and similarly, the infected individuals can be sorted into different classes according to their infectivity. Moreover, some diseases may develop through several stages. Many authors have shown that the individuals＇ relation can be viewed as a complex network. So in this paper, in order to better explain the dynamical behavior of epidemics, we consider different epidemic models on complex networks, and obtain the epidemic threshold for each ease. Finally, we present numerical simulations for each case to verify our results.     

A Formula of Average Path Length for Unweighted Networks

In this paper, based on the adjacency matrix of the network and its powers, the formulas are derived for the shortest path and the average path length, and an effective algorithm is presented. Furthermore, an example is provided to demonstrate the proposed method.     

Research on Community Structure in Bus Transport Networks

We abstract the bus transport networks （BTNs） to two kinds of complex networks with space L and space P methods respectively. Using improved community detecting algorithm （PKM agglomerative algorithm）, we analyze the community property of two kinds of BTNs graphs. The results show that the BTNs graph described with space L method have obvious community property, but the other kind of BTNs graph described with space P method have not. The reason is that the BTNs graph described with space P method have the intense overlapping community property and general community division algorithms can not identify this kind of community structure. To overcome this problem, we propose a novel community structure called N-depth community and present a corresponding community detecting algorithm, which can detect overlapping community. Applying the novel community structure and detecting algorithm to a BTN evolution model described with space P, whose network property agrees well with real BTNs＇, we get obvious community property.