Retrieval and chaos in extremely dilutedQ-Ising neural networks

Using a probabilistic approach, the deterministic and the stochastic parallel dynamics of aQ-Ising neural network are studied at finiteQ and in the limitQ. Exact evolution equations are presented for the first time-step. These formulas constitute recursion relations for the parallel dynamics of the extremely diluted asymmetric versions of these networks. An explicit analysis of the retrieval properties is carried out in terms of the gain parameter, the loading capacity, and the temperature. The results for theQ network are compared with those for theQ=3 andQ=4 models. Possible chaotic microscopic behavior is studied using the time evolution of the distance between two network configurations. For arbitrary finiteQ the retrieval regime is always chaotic. In the limitQ the network exhibits a dynamical transition toward chaos.     

Retrieval and chaos in layeredQ-Ising neural networks

Using a probabilistic approach, we study the parallel dynamics of theQ-Ising layered networks for arbitraryQ. By introducing auxiliary thermal fields, we express the stochastic dynamics within the gain function formulation of the deterministic dynamics. Evolution equations are derived for arbitraryQ at both zero and finite temperatures. An explicit analysis of the fixed-point equations is carried out for bothQ=3 andQ. The retrieval properties are discussed in terms of the gain parameter, the storage capacity, and the temperature. Using the time evolution of the distance between two network configurations, we investigate the possibility of microscopic chaos. Chaotic behavior is always present for arbitrary finiteQ. However, in the limitQ the existence of chaos depends on the parameters of the system.     

Chaotic dynamics of high-order neural networks

The dynamics of an extremely diluted neural network with high-order synapses acting as corrections to the Hopfield model is investigated. The learning rules for the high-order connections contain mixing of memories, different from all the previous generalizations of the Hopfield model. The dynamics may display fixed points or periodic and chaotic orbits, depending on the weight of the high-order connections , the noise levelT, and the network load, defined as the ratio between the number of stored patterns and the mean connectivity per neuron, =P/C. As in the related fully connected case, there is an optimal value of the weight that improves the storage capacity of the system (the capacity diverges).     

On the parallel dynamics for the Little-Hopfield model

We propose a method (algorithm) for calculation of the explicit formulas for evolution of the main and the residual overlaps. It allows us to confirm the Gardner-Derrida-Mottishaw second-step formula for the main overlap and to go beyond to the next steps. We discuss the dynamical status of the Amit-Gutfreund-Sompolinsky formula for the main overlap and some computersimulation results.     

Theory of the dynamics of the Hopfield model of associative memory

We present an analysis of the parallel dynamics of the Hopfield model of the associative memory of a neural network without recourse to the replica formalism. A probabilistic method based on the signal-to-noise ratio is employed to obtain a simple recursion relation for the zero temperature as well as the finite temperature dynamics of the network. The fixed points of the recursion relation and their basins of attraction are found to be in fairly satisfactory agreement with the numerical simulations of the model. We also present some new numerical results which support our recursion relation and throw light on the nature of the ensemble of the network states which are optimized with respect to single spin flips.     

Stochastic analyses of the dynamics of generalized Little-Hopfield-Hemmen type neural networks

Stochastic analyses are conducted of model neural networks of the generalized Little-Hopfield-Hemmen type, in which the synaptic connections with linearly embeddedp sets of patterns are free of symmetric ones, and a Glauber dynamics of a Markovian type is assumed. Two kinds of approaches are taken to study the stochastic dynamical behavior of the network system. First, by developing the method of the nonlinear master equation in the thermodynamic limitN, an exact self-consistent equation is derived for the time evolultion of the pattern overlaps which play the role of the order parameters of the system. The self-consistent equation is shown to describe almost completely the macroscopic dynamical behavior of the network system. Second, conducting the system-size expansion of the master equation for theN-body probability distribution of the Glauber dynamics makes it possible to analyze the fluctuations. In the course of the analysis, the self-consistent equation for the pattern overlaps is derived again. The main result of the rigorous fluctuation analysis is that as far as the fluctuations are concerned, the time course of the pattern overlap fluctuations behaves independently of the fluctuations in the remaining modes of the system's macrovariables, in accordance with the self-determining property of the macroscopic motion of the pattern overlaps for neural networks with linear synaptic couplings.     

Hybrid phase transitions of spreading dynamics in multiplex networks

In this paper, we study the spreading dynamics of social behaviors and focus on heterogenous responses of individuals depending on whether they realize the spreading or not. We model the system with a two-layer multiplex network, in which one layer describes the spreading of social behaviors and the other layer describes the diffusion of the awareness about the spreading. We use the susceptible-infected-susceptible (SIS) model to describe the dynamics of an individual if it is unaware of the spreading of the behavior. While when an individual is aware of the spreading of the social behavior its dynamics will follow the threshold model, in which an individual will adopt a behavior only when the fraction of its neighbors who have adopted the behavior is above a certain threshold. We find that such heterogenous reactions can induce intriguing dynamical properties. The dynamics of the whole network may exhibit hybrid phase transitions with the coexistence of continuous phase transition and bi-stable states. Detailed study of how the diffusion of the awareness influences the spreading dynamics of social behavior is provided. The results are supported by theoretical analysis.     

Detecting the structure of complex networks by quantum bosonic dynamics

In this paper, we introduce a non-interacting boson model to investigate the topological structure of complex networks. By exactly solving this model, we show that it provides a powerful analytical tool in uncovering the important properties of realistic networks. We find that the ground-state degeneracy of this model is equal to the number of connected components in a network and the square of each coefficient in the expansion of the ground state gives the average time that a random walker spends at each node in the infinite time limit. To show the usefulness of this approach in practice, we also carry out numerical simulations on some concrete complex networks. Our results are completely consistent with the previous conclusions derived by graph theory methods. Furthermore, we show that the first excited state appears always on the largest connected component of the network. The relationship between the first excited energy and the average shortest path length in networks is also discussed.     

无标度网络上的传播动力学

介绍了无标度网络上的传播动力学,在susceptible-infected-susceptible（SIS）模型的基础上考察了一般情况下无标度网络中疾病爆发的临界点问题,得出了关于临界点一般性的表达式.得到的结果在特殊情况下分别退化为已有的一些经典结论.同时分别讨论了这些情况的建模意义和可靠性. 关键词： 无标度网络 传播动力学 susceptible-infected-susceptible模型 临界点     

Dominant vertices in regulatory networks dynamics

Discrete-time regulatory networks are dynamical systems on directed graphs with a structure that is inspired on natural systems of interacting units. Using a notion of determination between vertices, we define sets of dominant vertices, and we prove that in the asymptotic regime, the projection of the dynamics on a dominant set allows us to determine the state of the whole system at all times. We provide an algorithm to find sets of dominant vertices, and we test its accuracy on several examples. We also explore the possibility of using the dominant set characteristics as a measure of the structural complexity of networks.     

Topology and dynamics of attractor neural networks: The role of loopiness

We derive an exact representation of the topological effect on the dynamics of sequence processing neural networks within signal-to-noise analysis. A new network structure parameter, loopiness coefficient, is introduced to quantitatively study the loop effect on network dynamics. A large loopiness coefficient means a high probability of finding loops in the networks. We develop recursive equations for the overlap parameters of neural networks in terms of their loopiness. It was found that a large loopiness increases the correlation among the network states at different times and eventually reduces the performance of neural networks. The theory is applied to several network topological structures, including fully-connected, densely-connected random, densely-connected regular and densely-connected small-world, where encouraging results are obtained.     

Analyzing neuronal networks using discrete-time dynamics

We develop mathematical techniques for analyzing detailed Hodgkin-Huxley like models for excitatory-inhibitory neuronal networks. Our strategy for studying a given network is to first reduce it to a discrete-time dynamical system. The discrete model is considerably easier to analyze, both mathematically and computationally, and parameters in the discrete model correspond directly to parameters in the original system of differential equations. While these networks arise in many important applications, a primary focus of this paper is to better understand mechanisms that underlie temporally dynamic responses in early processing of olfactory sensory information. The models presented here exhibit several properties that have been described for olfactory codes in an insect’s Antennal Lobe. These include transient patterns of synchronization and decorrelation of sensory inputs. By reducing the model to a discrete system, we are able to systematically study how properties of the dynamics, including the complex structure of the transients and attractors, depend on factors related to connectivity and the intrinsic and synaptic properties of cells within the network.     

具有时滞的一阶细胞神经网络动态行为研究

研究了具有时滞的一阶细胞神经网络的复杂动力学行为-证明了Hopf分叉的存在性，指出了 若选择适当的参数，则该网络中可以产生混沌-计算机仿真实验表明所得结论的正确性-由于 该网络的构成较简单，因此实际中可以作为混沌发生器加以应用- 关键词： 细胞神经网络 时滞 Hopf分叉 混沌     

Parallel Dynamics of Fully Connected Q-Ising Neural Networks

Using a probabilistic approach, the parallel dynamics of fully connected Q-Ising neural networks is studied for arbitrary Q. A novel recursive scheme is set up to determine the time evolution of the order parameters through the evolution of the distribution of the local field, taking into account all feedback correlations. In contrast to extremely diluted and layered network architectures, the local field is no longer normally distributed but contains a discrete part. As an illustrative example, an explicit analysis is carried out for the first four time steps. For the case of the Q = 2 and Q = 3 model the results are compared with extensive numerical simulations and excellent agreement is found. Finally, equilibrium fixed-point equations are derived and compared with the thermodynamic approach based upon the replica-symmetric mean-field approximation.     

MDSLB:A new static load balancing method for parallel molecular dynamics simulations

Large-scale parallelization of molecular dynamics simulations is facing challenges which seriously affect the simula- tion efficiency, among which the load imbalance problem is the most critical. In this paper, we propose, a new molecular dynamics static load balancing method （MDSLB）. By analyzing the characteristics of the short-range force of molecular dynamics programs running in parallel, we divide the short-range force into three kinds of force models, and then pack- age the computations of each force model into many tiny computational units called ＂cell loads＂, which provide the basic data structures for our load balancing method. In MDSLB, the spatial region is separated into sub-regions called ＂local domains＂, and the cell loads of each local domain are allocated to every processor in turn. Compared with the dynamic load balancing method, MDSLB can guarantee load balance by executing the algorithm only once at program startup without migrating the loads dynamically. We implement MDSLB in OpenFOAM software and test it on TianHe-lA supercomputer with 16 to 512 processors. Experimental results show that MDSLB can save 34%-64% time for the load imbalanced cases.     

Macroscopic Lyapunov functions for separable stochastic neural networks with detailed balance

We derive macroscopic Lyapunov functions for large, long-range, Ising-spin neural networks with separable symmetric interactions, which evolve in time according to local field alignment. We generalize existing constructions, which correspond todeterministic (zero-temperature) evolution and to specific choices of the interaction structure, to the case ofstochastic evolution and arbitrary separable interaction matrices, for both parallel and sequential spin updating. We find a direct relation between the form of the Lyapunov functions (which describe dynamical processes) and the saddle-point integration that results from performing equilibrium statistical mechanical studies of the present type of model.     

On training feedforward neural networks

A new algorithm that mapsn-dimensional binary vectors intom-dimensional binary vectors using 3-layered feedforward neural networks is described. The algorithm is based on a representation of the mapping in terms of the corners of then-dimensional signal cube. The weights to the hidden layer are found by a corner classification algorithm and the weights to the output layer are all equal to 1. Two corner classification algorithms are described. The first one is based on the perceptron algorithm and it performs generalization. The computing power of this algorithm may be gauged from the example that the exclusive-Or problem that requires several thousand iterative steps using the backpropagation algorithm was solved in 8 steps. Another corner classification algorithm presented in this paper does not require any computations to find the weights. However, in its basic form it does not perform generalization.     

Parallel Dynamics of Extremely Diluted Symmetric Q-Ising Neural Networks

The parallel dynamics of extremely diluted symmetric Q-Ising neural networks is studied for arbitrary Q using a probabilistic approach. In spite of the extremely diluted architecture the feedback correlations arising from the symmetry prevent a closed-form solution in contrast with the extremely diluted asymmetric model. A recursive scheme is found determining the complete time evolution of the order parameters taking into account all feedback. It is based upon the evolution of the distribution of the local field, as in the fully connected model. As an illustrative example an explicit analysis is carried out for the Q=2 and Q=3 model. These results agree with and extend the partial results existing for Q=2. For Q>2 the analysis is entirely new. Finally, equilibrium fixed-point equations are derived and a capacity-gain function diagram is obtained.