Multi-Layers Chaotic Neural Networks

Yf Hao,X Sun,Js Wang

2001-01-01

Abstract:Human brain can be regarded as a complicated multi-layers dynamical system which including plenty of chaotic neural groups. In this paper, a simplified multi-layers chaotic neural network is proposed and found that its dynamical characteristics are more complexity than those in a neural group through computer simulation.

What problem does this paper attempt to address?

Multi-frequency Sinusoidal Chaotic Neural Network and Its Complex Dynamics

Li Ru-Yi,Wang Guang-Yi,Dong Yu-Jiao,Zhou Wei

DOI: https://doi.org/10.7498/aps.69.20200725

2020-01-01

Abstract:A large number of animal experiments show that there is irregular chaos in the biological nervous systems. An artificial chaotic neural network is a highly nonlinear dynamic system, which can realize a series of complex dynamic behaviors, optimize global search and neural computation, and generate pseudo-random sequences for information encryption. According to the superposition theory of sinusoidal signals with different frequencies of brain waves, a non-monotone activation function based on the multifrequency-frequency conversion sinusoidal function and a piecewise function is proposed to make a neural network more consistent with the biological characteristics. The analysis shows that by adjusting the parameters, the activation function can exhibit the EEG signals in its different states, which can simulate the rich and varying brain activities when the brain waves of different frequencies and types work at the same time. According to the activation function we design a new chaotic cellular neural network. The complexity of the chaotic neural network is analyzed by the structural complexity based SE algorithm and C0 algorithm. By means of Lyapunov exponential spectrum, bifurcation diagram and basin of attraction, the effects of the activation function’s parameters on its dynamic characteristics are analyzed in detail, and it is found that a series of complex phenomena appears in the chaotic neural network, such as many different types of chaotic attractors, coexistent chaotic attractors and coexistence limit cycles, which improves the performance of the chaotic neural network, and proves that the multi-frequency sinusoidal chaotic neural network has rich dynamic characteristics, so it has a good prospect in information processing, information encryption and other aspects.
Fuzzy Chaotic Neural Networks

Tang Mo,Wang Kejun,Zhang Jianmin,Zheng Liying

DOI: https://doi.org/10.4018/978-1-60566-310-4.ch024

2009-01-01

Abstract:An understanding of the human brain’s local function has improved in recent years. But the cognition of human brain’s working process as a whole is still obscure. Both fuzzy logic and dynamic chaos are internal features of the human brain. Therefore, to fuse artificial neural networks, fuzzy logic and dynamic chaos together to constitute fuzzy chaotic neural networks is a novel method. This chapter is focused on the new ways of fuzzy neural networks construction and its application based on the existing achievement in this field. Four types of fuzzy chaotic neural networks are introduced, namely chaotic recurrent fuzzy neural networks, cooperation fuzzy chaotic neural networks, fuzzy number chaotic neural networks and self-evolution fuzzy chaotic neural networks. Chaotic recurrent fuzzy neural networks model is developed based on existing recurrent fuzzy neural networks through introducing chaos mapping into the membership layer. As it is a dynamic system, the input of neuron not only processes the information of former monument but also contains chaos maps information which is provided by dynamic chaos. Cooperation fuzzy chaotic neural network is proposed on the basis of simplified T-S fuzzy chaotic neural networks and Aihara chaotic neuron. It realizes fuzzy reasoning process by a neural network structure in which the rule inference part is realized by chaotic neural networks. Then enlightened by fuzzy number neural networks we propose a fuzzy number chaotic neuron, which is obtained by blurring the Aihara chaotic neuron. Using these neurons to construct fuzzy number chaotic neural networks, the mathematical model and weight updating rules are also given. At last, a self-evolution fuzzy chaotic neural network is proposed according to the principle of self-evolution network, which unifies the fuzzy Hopfield neural network constitution method.
Controlling Chaos In A Neural Network Based On The Phase Space Constraint

Gg He,Zt Cao,Hp Chen,P Zhu

DOI: https://doi.org/10.1142/S0217979203022192

2003-01-01

International Journal of Modern Physics B

Abstract:The chaotic neural network constructed with chaotic neurons exhibits very rich dynamic behaviors and has a nonperiodic associative memory. In the chaotic neural network, however, it is difficult to distinguish the stored patters from others, because the states of output of the network are in chaos. In order to apply the nonperiodic associative memory into information search and pattern identification, etc, it is necessary to control chaos in this chaotic neural network. In this paper, the phase space constraint method focused on the chaotic neural network is proposed. By analyzing the orbital of the network in phase space, we chose a part of states to be disturbed. In this way, the evolutional spaces of the strange attractors are constrained. The computer simulation proves that the chaos in the chaotic neural network can be controlled with above method and the network can converge in one of its stored patterns or their reverses which has the smallest Hamming distance with the initial state of the network. The work clarifies the application prospect of the associative dynamics of the chaotic neural network.
A Chaotic Neural Network Model

JW Shuai,ZX Chen,RT Liu,BX Wu

DOI: https://doi.org/10.1088/0256-307x/13/3/008

1996-01-01

Chinese Physics Letters

Abstract:A chaotic neural network model is introduced in this letter. With the variation of a controlling parameter, the model can process the input by a content-addressable means or fall into a chaotic state. Hence the features of brain such as the clear active state and the chaotic rest state can be simply simulated.
Exploiting Chaotic Dynamics as Deep Neural Networks

Shuhong Liu,Nozomi Akashi,Qingyao Huang,Yasuo Kuniyoshi,Kohei Nakajima

2024-05-30

Abstract:Chaos presents complex dynamics arising from nonlinearity and a sensitivity to initial states. These characteristics suggest a depth of expressivity that underscores their potential for advanced computational applications. However, strategies to effectively exploit chaotic dynamics for information processing have largely remained elusive. In this study, we reveal that the essence of chaos can be found in various state-of-the-art deep neural networks. Drawing inspiration from this revelation, we propose a novel method that directly leverages chaotic dynamics for deep learning architectures. Our approach is systematically evaluated across distinct chaotic systems. In all instances, our framework presents superior results to conventional deep neural networks in terms of accuracy, convergence speed, and efficiency. Furthermore, we found an active role of transient chaos formation in our scheme. Collectively, this study offers a new path for the integration of chaos, which has long been overlooked in information processing, and provides insights into the prospective fusion of chaotic dynamics within the domains of machine learning and neuromorphic computation.

Neural and Evolutionary Computing,Machine Learning
A Chaotic Neural Network and Its Application for Optimization

徐耀群,郝燕玲,孙枫,甲继承

DOI: https://doi.org/10.3969/j.issn.1672-0946.2001.04.007

2001-01-01

Abstract:This paper presents a kind of chaotic neural network,which combines stochastic with deterministic property to introduce chaos mechanism into Hopfield neural network to cursory search the optimum under chaotic dynamics and control the chaotic dynamics by an-nealing strategy to perform inverse bifurcation and disappear.After that,the gradient property of HNN is used to reach stable point.Simulation results show that such an algorithm can avoid getting stuck in local minima.
Controlling a chaotic neural network for information processing

Yang Li,Ping Zhu,Xiaoping Xie,Hongping Chen,Kazuyuki Aihara,Guoguang He

DOI: https://doi.org/10.1016/j.neucom.2012.11.024

IF: 6

2013-01-01

Neurocomputing

Abstract:A dynamic phase-space constraint method is proposed to control complex chaotic dynamics in a chaotic neural network (CNN), by limiting refractoriness internal states with a time-varying threshold. The limiting threshold evolves according to a control signal derived from the feedback internal states of the network. Simulation results reveal that the CNN under control exhibits multiphase behavior in the control parameter space. With proper parameter values, the controlled CNN converges to a periodic orbit which includes a stored pattern that has the smallest Hamming distance to its initial state. The properties of the controlled CNN can be used for information processing such as memory retrieval and pattern recognition.
Multi-layer Neural Network Involving Chaos Neurons and Its Application to Traffic-flow Prediction

DONG Chao-jun,LIU Zhi-yong

DOI: https://doi.org/10.3969/j.issn.1004-731x.2007.19.023

2007-01-01

Abstract:To investigate multi-layer chaos neural networks (MLCNN) and its application in traffic-flow forecasting,a multi-layer chaos network involving chaos neurons in hidden layer was developed based on BP network and chaos theory. The learning ability of the MLCNN was evaluated by benchmark experiment with XOR problem,and the experimental results show the feasibility of the chaos neurons to reinforce the learning ability and non-linear of the network. Urban traffic system could be regard as a chaotic system,so the MLCNN was used in prediction to traffic flow. The test result to certain intersection in Jiangmen city shows that the prediction error could be controlled below (or around) 10%. The MLCNN could also be used in prediction and control to the other chaotic systems.
Chaotic Neural Networks and Their Applications

YY He,LP Wang

DOI: https://doi.org/10.1109/wcica.2000.863345

2000-01-01

Abstract:Many difficult combinatorial optimization problems arising from science and technology are often difficult to solve exactly. Hence a great number of approximate algorithms for solving combinatorial optimization problems have been developed (Reeves, 1993, Zanakis et al., 1989). Hopfield and Tank (1985) applied the continuous-time, continuous-output Hopfield neural network (CTCO-HNN) to TSP, thereby initiating a new approach to optimization problems. But the Hopfield neural network is often trapped in local minima because of its gradient descent property. A number of modifications have been done on Hopfield neural networks for escaping from local minima. So far, incorporating chaos into the Hopfield neural network has been moved to be a successful approach to improving the convergent property of the HNNs. In this paper, we first review three chaotic neural network models, and then propose a novel approach to chaotic simulated annealing. Second, we apply all of them to a 10 city TSP. The time evolutions of energy functions and outputs of neurons for each model are given. The features and effectiveness of four methods are discussed and evaluated according to the simulation results. We conclude that the proposed neural network with simulated annealing has more powerful ability to obtain global minima than any other chaotic neural network model when applied to difficult combinatorial optimization problems
Adaptive chaotic controlling method of a chaotic neural network model

Lidan Wang,Shukai Duan,Guangyuan Liu

DOI: https://doi.org/10.1007/11427391_57

2005-01-01

Abstract:It has been found that chaotic dynamics may exist in real brain neurons and play important roles in signal proceeding. But it is hard to set suitable parameters of system to make it be chaotic in practice. In this paper, a general adaptive controlling method of nonlinear systems with chaotic dynamics is studied. According analysis Lyapunov exponent, the effectiveness of our scheme is illustrated by a series of computer simulations.
Optimal Path Selection In Computer Networks Using Chaotic Neural Networks

Huidang Zhang,Yuyao He

DOI: https://doi.org/10.1109/ICCIAS.2006.294172

2006-01-01

Abstract:This paper presents a neural network with chaotic dynamics to solve the optimal routing with the reduction of packet loss in computer network. The proposed chaotic neural network (CNN) can control network energy to increase, decrease or keep unchanged through chaotic controlled quantities added to each neuron, which can help neural network to enlarge searching space to get optimal solutions and avoid local mininia or invalid solutions. The cost function is also defined to represent the cost of optimal path with the reduction of packet loss. In order to verify the effectiveness, the optimal path problem is mapped onto a CNN of two dimensions and then 15-node computer network is optimized for path selection. From the experimental results, the success rate of obtaining optimal solutions of the proposed CNN are higher (3 % to 4 %) than that of GSNN, and much better (8 %to 14 %) than that of TCNN.
Research on Simulation and Modeling of Chaotic Neural Networks

NIE Yi-wen,SHI Wei-feng,HUANG Jian,LI Jun-jun

DOI: https://doi.org/10.3969/j.issn.1673-1255.2005.02.012

2005-01-01

Abstract:This paper describes basic principles of chaotic neural networks,and analyzes the relationship between the chaotic neural networks and nonlinear modeling.A general idea and a basic method for modeling with neural networks are put forward on the basis of the study of marine synchronous generator modeling based on chaotic neural networks.An example given in this paper shows that the application of the chaotic neural networks to the large nonlinear system simulation and modeling is a great success.
Complex Dynamical Behavior of Neural Networks in Circuit Implementation

Tiancai Wang,Xing He,Tingwen Huang

DOI: https://doi.org/10.1016/j.neucom.2016.01.030

IF: 6

2016-01-01

Neurocomputing

Abstract:Neural networks have abundant dynamic behaviors in nonlinear fields. In this paper, several circuit units are combined to realize some neuron-network models. The results of circuit simulation described by Multisim software demonstrates complex dynamical behaviors. It is shown that the phenomenon such as hyperchaos, limit cycles, homoclinic orbits of the designed circuits are closely similar to the results of numerical experiments.
Determining Topology Architecture for Chaotic Time Series Neural Network

Li Qing,Zheng Deling

DOI: https://doi.org/10.3321/j.issn:1001-053x.1999.01.026

1999-01-01

Abstract:Three-layered feed-forward neural network is used to establish the model for chaotictime series dynamic systems. Determination method of topology architecture for network is given.The size of the input node and hidden node is determined by the minimization the generalizationerror of the network. The simulation results show that the network architecture is optimized and thegeneralization error is dramatically decreased when this method is used.
Multilayer Perceptron Network Optimization for Chaotic Time Series Modeling.

Mu Qiao,Yanchun Liang,Adriano Tavares,Xiaohu Shi

DOI: https://doi.org/10.3390/e25070973

IF: 2.738

2023-01-01

Entropy

Abstract:Chaotic time series are widely present in practice, but due to their characteristics-such as internal randomness, nonlinearity, and long-term unpredictability-it is difficult to achieve high-precision intermediate or long-term predictions. Multi-layer perceptron (MLP) networks are an effective tool for chaotic time series modeling. Focusing on chaotic time series modeling, this paper presents a generalized degree of freedom approximation method of MLP. We then obtain its Akachi information criterion, which is designed as the loss function for training, hence developing an overall framework for chaotic time series analysis, including phase space reconstruction, model training, and model selection. To verify the effectiveness of the proposed method, it is applied to two artificial chaotic time series and two real-world chaotic time series. The numerical results show that the proposed optimized method is effective to obtain the best model from a group of candidates. Moreover, the optimized models perform very well in multi-step prediction tasks.
Study on Chaotic Prediction for Multi-branch Time Delay Neural Network

XI Jian-hui,HAN Min

DOI: https://doi.org/10.3969/j.issn.1002-0411.2007.02.010

2007-01-01

Information and Control

Abstract:A new multi-branch time delay neural network is adopted to conduct prediction research on chaotic time series.In the cases that the initial network states are not strictly equal to those of the practical systems,the approximation ability of this network to nonlinear system is discussed.The structure of the network is defined by integrating the theory of reconstructing phase space,which makes the efficient prediction information be contained in the network.The chaotic time series generated by Rossler chaotic equation and the practically observed yearly sunspot time series are respectively taken as examples.Simulations show that the presented network can be used to model and predict the chaos system successfully,and the method can get a higher precision.
Chaos Algorithm of Multilayer Feedforward Neural Network

Yaoqun Xu,Wang Shufu,Hao Yanling,Feng Sun

DOI: https://doi.org/10.1109/wcica.2000.863352

2002-01-01

Abstract:By the use of the properties of ergodicity, a chaos algorithm for a multilayer feedforward neural network is proposed with combined logistic mapping and BP algorithm. This algorithm can make the BP algorithm skip the local minimum and can find the global minimum finally.
The Multiple Frequency Conversion Sinusoidal Chaotic Neural Network and Its Application

Zhiqiang Hu,Zhongjin Guo,Gongming Wang,Lei Wang,Xiaodong Zhao,Yongfeng Zhang

DOI: https://doi.org/10.3390/fractalfract7090697

IF: 3.577

2023-01-01

Fractal and Fractional

Abstract:Aiming at the problem that the global search performance of a transiently chaotic neural network is not ideal, a multiple frequency conversion sinusoidal chaotic neural network (MFCSCNN) model is proposed based on the biological mechanism of the brain, including multiple functional modules and sinusoidal signals of different frequencies. In this model, multiple FCS functions and Sigmoid functions with different phase angles were used to construct the excitation function of neurons in the form of weighted sum. In this paper, the inverted bifurcation diagram, Lyapunov exponential diagram and parameter range of the model are given. The dynamic characteristics of the model are analyzed and applied to function optimization and combinatorial optimization problems. Experimental results show that the multiple frequency conversion sinusoidal chaotic neural network has better global search performance than the transient chaotic neural network and other related models.
A Review of Chaotic Systems Based on Memristive Hopfield Neural Networks

Hairong Lin,Chunhua Wang,Fei Yu,Jingru Sun,Sichun Du,Zekun Deng,Quanli Deng

DOI: https://doi.org/10.3390/math11061369

IF: 2.4

2023-01-01

Mathematics

Abstract:Since the Lorenz chaotic system was discovered in 1963, the construction of chaotic systems with complex dynamics has been a research hotspot in the field of chaos. Recently, memristive Hopfield neural networks (MHNNs) offer great potential in the design of complex, chaotic systems because of their special network structures, hyperbolic tangent activation function, and memory property. Many chaotic systems based on MHNNs have been proposed and exhibit various complex dynamical behaviors, including hyperchaos, coexisting attractors, multistability, extreme multistability, multi-scroll attractors, multi-structure attractors, and initial-offset coexisting behaviors. A comprehensive review of the MHNN-based chaotic systems has become an urgent requirement. In this review, we first briefly introduce the basic knowledge of the Hopfiled neural network, memristor, and chaotic dynamics. Then, different modeling methods of the MHNN-based chaotic systems are analyzed and discussed. Concurrently, the pioneering works and some recent important papers related to MHNN-based chaotic systems are reviewed in detail. Finally, we survey the progress of MHNN-based chaotic systems for application in various scenarios. Some open problems and visions for the future in this field are presented. We attempt to provide a reference and a resource for both chaos researchers and those outside the field who hope to apply chaotic systems in a particular application.
A Model of Chaotic Neural Network with Applications in Optimization

张强,马润年,许进

DOI: https://doi.org/10.3321/j.issn:1001-506x.2002.02.016

2002-01-01

Abstract:A neural network model with transient chaotic dynamic behaviors is proposed by introducing a nonlinear self-feedback into canonical Hopfield neural networks(HNN). The model gradually approaches,through reversed period-doubling bifurcations, to a dynamical structure similar to the Hopfield neural network which converges to a stable equilibrium point. As the model has rich dynamics, it can be expected to have robust searching ability for globally optimal solutions. Finally, two examples of function optimization are given to show the validity of this model.

Multi-Layers Chaotic Neural Networks

Multi-frequency Sinusoidal Chaotic Neural Network and Its Complex Dynamics

Fuzzy Chaotic Neural Networks

Controlling Chaos In A Neural Network Based On The Phase Space Constraint

A Chaotic Neural Network Model

Exploiting Chaotic Dynamics as Deep Neural Networks

A Chaotic Neural Network and Its Application for Optimization

Controlling a chaotic neural network for information processing

Multi-layer Neural Network Involving Chaos Neurons and Its Application to Traffic-flow Prediction

Chaotic Neural Networks and Their Applications

Adaptive chaotic controlling method of a chaotic neural network model

Optimal Path Selection In Computer Networks Using Chaotic Neural Networks

Research on Simulation and Modeling of Chaotic Neural Networks

Complex Dynamical Behavior of Neural Networks in Circuit Implementation

Determining Topology Architecture for Chaotic Time Series Neural Network

Multilayer Perceptron Network Optimization for Chaotic Time Series Modeling.

Study on Chaotic Prediction for Multi-branch Time Delay Neural Network

Chaos Algorithm of Multilayer Feedforward Neural Network

The Multiple Frequency Conversion Sinusoidal Chaotic Neural Network and Its Application

A Review of Chaotic Systems Based on Memristive Hopfield Neural Networks

A Model of Chaotic Neural Network with Applications in Optimization