Flow invariance for competitive neural networks with different time-scales

This paper deals with the adaptive nonlinear identification and trajectory tracking via dynamic multilayer neural network (NN) with different timescales. Two NN identifiers are proposed for nonlinear systems identification via dynamic NNs with different timescales including both fast and slow phenomenon. The first NN identifier uses the output signals from the actual system for the system identification. In the second NN identifier, all the output signals from nonlinear system are replaced with the state variables of the NNs. The online identification algorithms for both NN identifier parameters are proposed using Lyapunov function and singularly perturbed techniques. With the identified NN models, two indirect adaptive NN controllers for the nonlinear systems containing slow and fast dynamic processes are developed. For both developed adaptive NN controllers, the trajectory errors are analyzed and the stability of the systems is proved. Simulation results show that the controller based on the second identifier has better performance than that of the first identifier.

The dynamics of cortical cognitive maps developed by selforganization must include the aspects of long and short-term memory. The behavior of such a neural network is characterized by an equation of neural activity as a fast phenomenon and an equation of synaptic modification as a slow part of the neural system, besides, this model bases on unsupervised synaptic learning algorithm. In this paper, using theory of the topological degree and strict Liapunov functional methods, we prove existence and uniqueness of the equilibrium of competitive neural networks with continuously distributed delays and different time scales, and present some new criteria for its global robust stability.

Complex networks are ubiquitous even in the form of defense networks in many fields such as ecological immune systems, biological immune systems, brain networks, air defense systems, national security, cyber security, peace and conflict studies, and so on. In the real world, complex networks exist in both static and dynamic forms. By understanding probabilistic failures and their possible defense mechanisms in static and dynamic complex networks, scientists and researchers can make decisions in developing and implementing methodologies of many fields as we can say, from the fields of biology to fields of technology. In this research, static and dynamic scale-free and small world networks have been generated as complex networks with network generation algorithms. Besides that, probabilistic failures and their possible defense mechanisms have also been generated and designed for previously generated complex networks. On these generated probabilistic failures and their possible defense mechanism, several mechanistic and statistical analysis methods have been applied. To generate these static and dynamic complex networks and their probabilistic failures and possible defense mechanism, two modified network generation algorithms along with several network measurements algorithms have been implemented where parameters have been chosen to produce possible types of static and dynamic scale-free and small world networks, and their corresponding probabilistic failures and possible defense mechanisms. Besides analyzing probabilistic failures and possible defense mechanisms for generated complex networks; the robustness, stability and sustainability of those networks have been studied and analyzed for probabilistic failures and their possible defense mechanisms.

Fault detection and diagnosis for non-linear processes empowered by dynamic neural networks

Several mathematical approaches to studying analytically the dynamics of neural networks rely on mean-field approximations, which are rigorously applicable only to networks of infinite size. However, all existing real biological networks have finite size, and many of them, such as microscopic circuits in invertebrates, are composed only of a few tens of neurons. Thus, it is important to be able to extend to small-size networks our ability to study analytically neural dynamics. Analytical solutions of the dynamics of small-size neural networks have remained elusive for many decades, because the powerful methods of statistical analysis, such as the central limit theorem and the law of large numbers, do not apply to small networks. In this article, we critically review recent progress on the study of the dynamics of small networks composed of binary neurons. In particular, we review the mathematical techniques we developed for studying the bifurcations of the network dynamics, the dualism between neural activity and membrane potentials, cross-neuron correlations, and pattern storage in stochastic networks. Then, we compare our results with existing mathematical techniques for studying networks composed of a finite number of neurons. Finally, we highlight key challenges that remain open, future directions for further progress, and possible implications of our results for neuroscience.

Complex dynamical networks consist of a large set of interconnected nodes with each node being a fundamental unit with detailed contents. A great number of natural and man-made complex networks such as social networks, food networks, neural networks, the World Wide Web, computer networks, electrical power grid and so on, can be effectively modelled by complex dynamical networks. Synchronization of a group of dynamical nodes in a complex network topology is one of the most interesting and significant phenomenon in complex dynamical networks. In this section, we shall discuss impulsive synchronization of the discrete delay networks. By employing the impulsive control method and the Razumikhin-type results obtained in previous chapter, two criteria for exponential synchronization will be derived.

Information dissemination has become one of the most important services of communication networks. Modeling the diffusion of information through such networks is crucial for our modern information societies. In this work, novel models, segregating between useful and malicious types of information, are introduced, in order to better study Information Dissemination Dynamics (IDD) in wireless complex communication networks, and eventually allow taking into account special network features in IDD. According to the proposed models, and inspired from epidemiology, we investigate the IDD in various complex network types through the use of the Susceptible-Infected (SI) paradigm for useful information dissemination and the Susceptible-Infected-Susceptible (SIS) paradigm for malicious information spreading. We provide analysis and simulation results for both types of diffused information, in order to identify performance and robustness potentials for each dissemination process with respect to the characteristics of the underlying complex networking infrastructures. We demonstrate that the proposed approach can generically characterize IDD in wireless complex networks and reveal salient features of dissemination dynamics in each network type, which could eventually aid in the design of more advanced, robust, and efficient networks and services.

In this chapter, we provide a basic introduction to complex and adaptive dynamical networks. We introduce graph theoretical preliminaries along with different classes of networks studied in the subsequent chapters. We discuss different types of models and coupling schemes. Further, we show how adaptivity is modeled in the context of synaptic plasticity and outline the reduction to phase oscillator models with phase difference-dependent plasticity.

Complex systems provide an opportunity to analyze the essence of phenomena by studying their intricate connections. The networks formed by these connections, known as complex networks, embody the underlying principles governing the system’s behavior. While complex networks have been previously applied in the field of evolutionary computation, prior studies have been limited in their ability to reach conclusive conclusions. Based on our investigations, we are against the notion that there is a direct link between the complex network structure of an algorithm and its performance, and we demonstrate this experimentally. In this paper, we address these limitations by analyzing the dynamic complex network structures of five algorithms across three different problems. By incorporating mathematical distributions utilized in prior research, we not only generate novel insights but also refine and challenge previous conclusions. Specifically, we introduce the biased Poisson distribution to describe the algorithm’s exploration capability and the biased power-law distribution to represent its exploitation potential during the convergence process. Our aim is to redirect research on the interplay between complex networks and evolutionary computation towards dynamic network structures, elucidating the essence of exploitation and exploration in the black-box optimization process of evolutionary algorithms via dynamic complex networks.

One of the difficulties encountered in the application of reinforcement learning methods to real-world problem is the generalization of large-scale or continuous state space. In order to solve the curse of dimensionality problem caused by discretizing continuous state space, a kind of Q-learning method for continuous state space based on a dynamic Elman neural network was proposed in this paper. The inputs and the output of Elman network are the system state-action pair and the corresponding Q-value. That is, Elman network is used to estimate the Q-value of state-action pair on-line. Eligibility trace for connecting weights is introduced by borrowing ideas from the eligibility trace mechanism of state in temporal difference algorithm to improve the learning speed of neural network. Computer simulations on mountain car control illustrate the performance and applicability of the proposed Q-learning scheme.

Community structure is an important feature of complex networks. These community structures have the fractal characteristics, that is, there is a self similarity of statistical sense between the complex networks and their local. There have been more and more recent researches on communities’ discovery in complex network. However, most existing approaches require the complete information of entire network, which is impractical for some networks, e.g. the dynamical network and the network that is too large to get the whole information. Therefore, the study of community discovery in complex networks has rather important theoretical and practical value. Through the analysis and study of the complex network evolution models with renormalization and the community change of the complex network evolution, using the tool of adjusting scales as the renormalization process, a multi-scale network community detection algorithm based on fractal feature evolution was proposed. The purpose is to solve community discovery problems in dynamic complex networks, and the effectiveness of the proposed method is verified by real data sets. By comparing result of this paper with the previous methods on some real world networks, and experimental results verify the feasibility and accuracy.

This manuscript explores the analysis of passivity and synchronization criteria for a complex fractional-order dynamical network model with derivative couplings and time-varying delays. The passivity problem of the proposed network model is deduced using various inequality methods and presented as a linear matrix inequality. To ensure complete synchronization for a fractional-order complex dynamical network with derivative couplings (CDNMDC), we derive suitable criteria using an adaptive feedback control method. Additionally, we investigate the synchronization criterion of these complex networks while accounting for parameter uncertainties. Finally, we provide an example to demonstrate the effectiveness of the proposed solutions.

An increasing number of applications of dynamic neural networks has been developed for digital signal processing (DSP), dynamic neural networks are feedforward neural networks with commonly used scalar synapses replaced by linear filters. This provides feedforward neural networks with the capability of performing dynamic mappings, which depend on past input values, dynamic neural networks are suitable for time series prediction, nonlinear system identification, and signal processing applications. Their most popular types are Finite Impulse Response (FIR) neural networks, which are obtained by replacing synapses with finite impulse response filters. Due to their guaranteed stability characteristic and easy to minimize error surface they have been used with great success in many applications such as signal enhancement, noise cancellation, classification of input patterns, system identification, prediction, and control.. Most of the works on system identification using neural networks are based on multilayer feedforward neural networks with backpropagation learning or more efficient variations of this algorithm, an elegant method for training layered networks. This paper is based on work in a Dynamic System Modeling (DYSMO) and as an application for speed control of DC motor drive.

The robustness of controlling complex networks is significant in network science. In this paper, we focus on evaluating and analyzing the robustness of controlling edge dynamics in complex networks against node failure. Using three categories of all nodes to quantify the robustness, we find that the percentages of the three types of nodes are mainly related to the degree distribution of networks. The simulation results of model networks and analytic calculations show that the sparse inhomogeneous networks, which emerge in many real complex networks, have strong control robustness from the point of the number of ordinary nodes, but the strong positive correlation between in and out degrees reduces the control robustness. Evaluation of real-world networks indicates that most of them have few or no critical nodes, that is, they do not need to increase driver nodes to maintain control for most of node failures. Then an adding circuit-link strategy is proposed to optimize the robustness of edge controllability.

The article is devoted to the problem of optimization of search request ranking algorithms in the digital information retrieval system. The algorithm of functioning of the neural network ranking unit based on Hopfield neural network is built. The ability to generate a ranked list of pages found as a result of the request in the digital information retrieval system can be provided by solving two problems of integer optimization: the problem of assignment of combinatorial sets of criteria for assessing the relevance of web page search and the problem of sorting of numbers—relevance values. The architecture of the neural network model based on the dynamic Hopfield neural network with binary output function designed for combinatorial optimization of the final list of documents found in the digital information retrieval system was synthesized. Promising variants of neural network models with binary output function of neurons for synthesis of the optimal evaluation plan with a combinatorial set of criteria by solving the problem of assignment were built. It has been proven that the built models differ in the rules for determining the coefficients of synaptic connections and external shifts; each of the created rules can be used independently or in different combinations with one another. In the course of analytical research, it was found that the optimization formulation of the problem of sorting of relevance values of search pages is identical to the problem of assignment of combinatorial groups of evaluation criteria provided that the elements of the performance matrix of the latter are defined as linear combinations of relevance values.