The essential order of approximation for neural networks

There have been various studies on approximation ability of feedforward neural networks (FNNs). Most of the existing studies are, however, only concerned with density or upper bound estimation on how a multivariate function can be approximated by an FNN, and consequently, the essential approximation ability of an FNN cannot be revealed. In this paper, by establishing both upper and lower bound estimations on approximation order, the essential approximation ability (namely, the essential approximation order) of a class of FNNs is clarified in terms of the modulus of smoothness of functions to be approximated. The involved FNNs can not only approximate any continuous or integrable functions defined on a compact set arbitrarily well, but also provide an explicit lower bound on the number of hidden units required. By making use of multivariate approximation tools, it is shown that when the functions to be approximated are Lipschitzian with order up to 2, the approximation speed of the FNNs is uniquely deter     

Global exponential stability in Lagrange sense for neutral type recurrent neural networks

In this paper, the global exponential stability in Lagrange sense for continuous neutral type recurrent neural networks (NRNNs) with multiple time delays is studied. Three different types of activation functions are considered, including general bounded and two types of sigmoid activation functions. By constructing appropriate Lyapunov functions, some easily verifiable criteria for the ultimate boundedness and global exponential attractivity of NRNNs are obtained. These results can be applied to monostable and multistable neural networks as well as chaos control and chaos synchronization.     

Global exponential stability of impulsive high-order Hopfield type neural networks with delays

In this paper, we investigate the global exponential stability of impulsive high-order Hopfield type neural networks with delays. By establishing the impulsive delay differential inequalities and using the Lyapunov method, two sufficient conditions that guarantee global exponential stability of these networks are given, and the exponential convergence rate is also obtained. A numerical example is given to demonstrate the validity of the results.     

Lower estimation of approximation rate for neural networks

Let SFd and Πψ,n,d = { nj=1bjψ(ωj·x+θj) :bj,θj∈R,ωj∈Rd} be the set of periodic and Lebesgue’s square-integrable functions and the set of feedforward neural network (FNN) functions, respectively. Denote by dist (SF d, Πψ,n,d) the deviation of the set SF d from the set Πψ,n,d. A main purpose of this paper is to estimate the deviation. In particular, based on the Fourier transforms and the theory of approximation, a lower estimation for dist (SFd, Πψ,n,d) is proved. That is, dist(SF d, Πψ,n,d) (nlogC2n)1/2 . T...     

有理式二阶前馈型神经网络

文章提出了二阶有理式多层前馈神经网络的数学模型。有理式多层神经网络的思想来源于函数逼近理论中的有理式逼近。有理式前馈神经网络模型是传统前俯神经网络模型的推广，能有效地求解函数逼近问题。文章给出了有理式多层神经网络的学习算法，即误差反传播学习算法。就计算复杂度而言，有理式神经网络的学习算法与传统的多层神经网络反传播算法是同阶的。文章还给出了函数逼近和模式识别两个应用实例，实验结果说明二阶有理式多层神经网络在解决传统的问题上是有效的。     

Global exponential convergence of periodic neural networks with time-varying delays

In this paper, the global exponential convergence of a general class of periodic neural networks with time-varying delays is investigated. Based on the theory of mixed monotone operator, a testable algebraic criteria for ascertaining global exponential convergence is derived. Furthermore, the rate of exponential convergence and bound of the networks are also estimated. Finally, a numerical example is given to show the effectiveness of the obtained results.     

Delay-dependent exponential state estimators for stochastic neural networks of neutral type with both discrete and distributed delays

This paper considered the state estimation for stochastic neural networks of neutral type with discrete and distributed delays. By using available output measurements, the state estimator can approximate the neuron states, and the asymptotic property of the state error is mean square exponential stable and also almost surely exponential stable in the presence of discrete and distributed delays. Under the Lipschitz assumptions for the activation functions and the measurement nonlinearity, a delay-dependent linear matrix inequality (LMI) criterion is proposed to guarantee the existence of the desired estimators by constructing an appropriate Lyapunov-Krasovskii function. It is shown that the existence conditions and the explicit expression of the state estimator can be parameterised in terms of the solution to a LMI. Finally, two numerical examples are presented to demonstrate the validity of the theoretical results and show that the theorem can provide less conservative conditions.     

Global exponential stability in Lagrange sense for periodic neural networks with various activation functions

In this paper, global exponential stability in Lagrange sense for periodic neural networks with various activation functions is further studied. By constructing appropriate Lyapunov-like functions, we provide easily verifiable criteria for the boundedness and global exponential attractivity of periodic neural networks. These theoretical analysis can narrow the search field of optimization computation, associative memories, chaos control and provide convenience for applications.     

快速二阶BP网络及其在城市用水量预测中的应用

针对BP网络收敛速度慢,易导致局部极小值的缺点,提出一种快速二阶BP网络,并以城市年用水量预测为例,与BP网络对比,结果表明,该方法加快了收敛速度,提出了结果的准确度。     

Global exponential stability analysis for cellular neural networks with variable coefficients and delays

Some sufficient conditions for the global exponential stability of cellular neural networks with variable coefficients and time-varying delays are obtained by a method based on a delayed differential inequality. The method, which does not make use of Lyapunov functionals, is simple and effective for the stability analysis of cellular neural networks with variable coefficients and time-varying delays. Some previous results in the literature are shown to be special cases of our results.      

Asymptotic stability of impulsive high-order Hopfield type neural networks

In this paper, we discuss impulsive high-order Hopfield type neural networks. Investigating their global asymptotic stability, by using Lyapunov function method, sufficient conditions that guarantee global asymptotic stability of networks are given. These criteria can be used to analyse the dynamics of biological neural systems or to design globally stable artificial neural networks. Two numerical examples are given to illustrate the effectiveness of the proposed method.     

Harmless delays for global exponential stability of Cohen–Grossberg neural networks

In this paper, the Cohen–Grossberg neural networks with time delays are considered without assuming any symmetry of connection matrix and differentiability of the activation functions. By constructing a novel Lyapunov functional, new sufficient criteria for the existence of a unique equilibrium and global exponential stability of the network are derived. These criteria are all independent of the magnitudes of delays, and so the delays under these conditions are harmless. Those results are shown to generalize the previous global exponential stability results derived in the literature.     

Global exponential synchronization of fuzzy cellular neural networks with delays and reaction-diffusion terms

The global exponential synchronization for a class of fuzzy cellular neural networks with delays and reaction-diffusion terms is discussed. Some new sufficient conditions are obtained by using the Lyapunov functional method, many real parameters and inequality techniques. The result is also easy to check and plays an important role in the design and application of globally exponentially synchronization. Finally, an example is given to verify our results.     

Global exponential stability analysis of delayed Cohen--Grossberg neural networks with distributed delays

In this article, the global exponential stability problem of Cohen--Grossberg neural networks with both discrete-time delays and distributed delays is investigated. The existence and global stability for the unique equilibrium of the Cohen--Grossberg neural networks with distributed delays are achieved by using some new Lyapunov functionals, M-matrix theory and some analytic techniques, and some less restrictive conditions are obtained. An example is also worked out to validate the advantages of our results.     

Global exponential stability for neutral-type BAM neural networks with time-varying delays

In this article, the global exponential stability of neutral-type bidirectional associative memory (BAM) neural networks with time-varying delays is analysed by utilizing the Lyapunov–Krasovskii functional and combining with the linear matrix inequality (LMI) approach. New sufficient conditions ensuring the global exponential stability of neutral-type BAM neural networks is obtained by using the powerful MATLAB LMI control toolbox. In addition, an example is provided to illustrate the applicability of the result.     

Global exponential stability results for delayed neural networks of neutral type

In this paper, by utilizing the Lyapunov–Krasovkii functional and combining with the linear-matrix inequality (LMI) approach, we analyse the global exponential stability of delayed neural networks of neutral type. In addition, the examples are provided to illustrate the applicability of the result using LMI control toolbox in MATLAB.     

Global exponential stability of Hopfield reaction-diffusion neural networks with time-varying delays

The authors analyze the existence of the equilibrium point and global exponential stability for Hopfield reaction-diffusion neural networks with time-varying delays by means of the topological degree theory and generalized Halanay inequality. Since the diffusion phenomena and time delay could not be ignored in neural networks and electric circuits, the model presented here is close to the actual systems, and the sufficient conditions on global exponential stability established in this paper, which are easily verifiable, have a wider adaptive range.     

Novel criteria for global robust exponential stability of neural networks with time-varying delays via LMI approach

The article addresses the problem of global robust exponential stability of interval neural networks with time-varying delays. On the basis of linear matrix inequality technique and M-matrix theory, some novel sufficient conditions for the existence, uniqueness, and global robust exponential stability of the equilibrium point for delayed interval neural networks are presented. It is shown that our results improve and generalize some previously published ones. Some numerical examples and simulations are given to show the effectiveness of the obtained results.     

Analysis of quantization effects on high-order function neural networks

In this paper we investigate the combined effects of quantization and clipping on high-order function neural networks (HOFNN). Statistical models are used to analyze the effects of quantization in a digital implementation. We analyze the performance degradation caused as a function of the number of fixed-point and floating-point quantization bits under the assumption of different probability distributions for the quantized variables, and then compare the training performance between situations with and without weight clipping. We establish and analyze the relationships for a true nonlinear neuron between inputs and outputs bit resolution, training and quantization methods, network order and performance degradation, all based on statistical models, and for on-chip and off-chip training. Our experimental simulation results verify the presented theoretical analysis.

Further results for global exponential stability of stochastic memristor-based neural networks with time-varying delays

In recent years, the stability problems of memristor-based neural networks have been studied extensively. This paper not only takes the unavoidable noise into consideration but also investigates the global exponential stability of stochastic memristor-based neural networks with time-varying delays. The obtained criteria are essentially new and complement previously known ones, which can be easily validated with the parameters of system itself. In addition, the study of the nonlinear dynamics for the addressed neural networks may be helpful in qualitative analysis for general stochastic systems. Finally, two numerical examples are provided to substantiate our results.