Fuzzy set-oriented neural networks based on fuzzy polynomial inference and dynamic genetic optimization

In this paper, we introduce a new topology and offer a comprehensive design methodology of fuzzy set-based neural networks (FsNNs). The proposed architecture of the FsNNs is based on the fuzzy polynomial neurons formed through a collection of ‘if-then’ fuzzy rules, fuzzy inference, and polynomials with the extended structure of the premise and the consequence parts of fuzzy rules being formed within the networks. Three different forms of regression polynomials (namely constant, linear, and quadratic) are used in the consequence part of the rules. In order to build an optimal FsNN, the underlying structural and parametric optimization is supported by a dynamic search-based genetic algorithm (GA), which forms an optimal solution through successive adjustments (refinements) of the search range. The structure optimization involves the determination of the input variables included in the premise part and the order of the polynomial forming the consequence part of the rules. In the study, we explore two types of optimization methodologies, namely a simultaneous tuning and a separate tuning. GAs are global optimizers; however, when being used in their generic version, they often lead to a significant computing overhead caused by the need to explore an excessively large search space. To eliminate this shortcoming and increase the effectiveness of the optimization itself, we introduce a dynamic search-based GA that results in a rapid convergence while narrowing down the search to a limited region of the search space. We exploit this optimization mechanism to be completed both at the structural as well as the parametric level. To evaluate the performance of the proposed FsNN, we offer a suite of several representative numerical examples.     

A fuzzy ensemble of parallel polynomial neural networks with information granules formed by fuzzy clustering

In this paper, we introduce a new category of fuzzy models called a fuzzy ensemble of parallel polynomial neural network (FEP²N²), which consist of a series of polynomial neural networks weighted by activation levels of information granules formed with the use of fuzzy clustering. The two underlying design mechanisms of the proposed networks rely on information granules resulting from the use of fuzzy C-means clustering (FCM) and take advantage of polynomial neural networks (PNNs).The resulting model comes in the form of parallel polynomial neural networks. In the design procedure, in order to estimate the optimal values of the coefficients of polynomial neural networks we use a weighted least square estimation algorithm. We incorporate various types of structures as the consequent part of the fuzzy model when using the learning algorithm. Among the diverse structures being available, we consider polynomial neural networks, which exhibit highly nonlinear characteristics when being viewed as local learning models.We use FCM to form information granules and to overcome the high dimensionality problem. We adopt PNNs to find the optimal local models, which can describe the relationship between the input variables and output variable within some local region of the input space.We show that the generalization capabilities as well as the approximation abilities of the proposed model are improved as a result of using polynomial neural networks. The performance of the network is quantified through experimentation in which we use a number of benchmarks already exploited within the realm of fuzzy or neurofuzzy modeling.     

多元多项式函数的三层前向神经网络逼近方法

该文首先用构造性方法证明:对任意r阶多元多项式,存在确定权值和确定隐元个数的三层前向神经网络.它能以任意精度逼近该多项式.其中权值由所给多元多项式的系数和激活函数确定,而隐元个数由r与输入变量维数确定.作者给出算法和算例,说明基于文中所构造的神经网络可非常高效地逼近多元多项式函数.具体化到一元多项式的情形,文中结果比曹飞龙等所提出的网络和算法更为简单、高效;所获结果对前向神经网络逼近多元多项式函数类的网络构造以及逼近等具有重要的理论与应用意义,为神经网络逼近任意函数的网络构造的理论与方法提供了一条途径.     

Use of neural networks in iterative learning control systems

This paper presents an approach to the use of neural networks to improve iterative learning control performance. The neural networks are used to estimate the learning gain of an iterative learning law and to store the learned control input profiles for different reference trajectories. A neural network of piecewise linear approximation is presented to identify effectively the system dynamics, and the approximation property and persistently exciting condition are discussed. In addition, training of a feedforward neural controller is presented to accumulate control information learned by an iterative update law for various reference trajectories. Then, an iterative learning law with a feedforward neural controller is suggested and its convergence property is stated with the convergence condition. The effectiveness of the present methods has been demonstrated through simulations by applying them to a two-link robot manipulator.     

基于RBF辨识的模糊神经网络控制器的设计与实现

随着众多新型模糊神经网络被提出,针对模糊神经网络具有的典型特点,即需要对输入输出数据范围进行转化和处理,所涉及到的对量化因子和比例因子的实时调节问题,该文提出一种优化方案。其依据神经网络具有的自学习能力,通过增加模糊神经网络的层数,提出一种包含对量化因子和比例因子调节的改进型模糊神经网络,以减少系统的辅助优化环节。同时,引入辨识性能较好的径向基函数神经网络（RBF）为系统提供精确的Jacobian信息,取代常规的近似做法。最后结合实例仿真证明了该优化方案的合理性。     

类矩阵神经核特征融合的人脸识别方法

针对人脸识别过程中所提取特征向量的信息不完整性与整体图像信息数据量较大的问题,提出一种类矩阵神经核特征融合的人脸识别方法。该方法为深度神经网络的首层升维操作,首先将人脸数据作为特征向量的集合,利用随机矩阵列采样构成随机特征矩阵;其次设计深度神经核将随机特征矩阵映射为高维空间中的新特征向量;最后利用快速收缩算法求解匹配过程中的不定线性代数方程组,使收敛速度达到二阶收敛。该方法既克服了直接使用人脸图像数据空间复杂度较大的问题,又增加了特征的非线性结构,提高了特征向量的表达能力。实验结果表明,该方法识别率高、稳定性强、鲁棒性好,适合处理大型数据。     

Fuzzy polynomial neural networks: hybrid architectures of fuzzy modeling

We introduce a concept of fuzzy polynomial neural networks (FPNNs), a hybrid modeling architecture combining polynomial neural networks (PNNs) and fuzzy neural networks (FNNs). The development of the FPNNs dwells on the technologies of computational intelligence (CI), namely fuzzy sets, neural networks, and genetic algorithms. The structure of the FPNN results from a synergistic usage of FNN and PNN. FNNs contribute to the formation of the premise part of the rule-based structure of the FPNN. The consequence part of the FPNN is designed using PNNs. The structure of the PNN is not fixed in advance as it usually takes place in the case of conventional neural networks, but becomes organized dynamically to meet the required approximation error. We exploit a group method of data handling (GMDH) to produce this dynamic topology of the network. The performance of the FPNN is quantified through experimentation that exploits standard data already used in fuzzy modeling. The obtained experimental results reveal that the proposed networks exhibit high accuracy and generalization capabilities in comparison to other similar fuzzy models.     

Neural Network Output Optimization Using Interval Analysis

The problem of output optimization within a specified input space of neural networks (NNs) with fixed weights is discussed in this paper. The problem is (highly) nonlinear when nonlinear activation functions are used. This global optimization problem is encountered in the reinforcement learning (RL) community. Interval analysis is applied to guarantee that all solutions are found to any degree of accuracy with guaranteed bounds. The major drawbacks of interval analysis, i.e., dependency effect and high-computational load, are both present for the problem of NN output optimization. Taylor models (TMs) are introduced to reduce these drawbacks. They have excellent convergence properties for small intervals. However, the dependency effect still remains and is even made worse when evaluating large input domains. As an alternative to TMs, a different form of polynomial inclusion functions, called the polynomial set (PS) method, is introduced. This new method has the property that the bounds on the network output are tighter or at least equal to those obtained through standard interval arithmetic (IA). Experiments show that the PS method outperforms the other methods for the NN output optimization problem.     

Hybrid fuzzy set-based polynomial neural networks and their development with the aid of genetic optimization and information granulation

We introduce a new architecture of feed-forward neural networks called hybrid fuzzy set-based polynomial neural networks (HFSPNNs) that are composed of heterogeneous feed-forward neural networks such as polynomial neural networks (PNNs) and fuzzy set-based polynomial neural networks (FSPNNs). We develop their comprehensive design methodology by embracing mechanisms of genetic optimization and information granulation. The construction of information granulation-driven HFSPNN exploits fundamental technologies of computational intelligence (CI), namely fuzzy sets, neural networks, and genetic algorithms (GAs). The architecture of the resulting information granulation-driven genetically optimized HFSPNN results from a synergistic usage of the hybrid system generated by combining original fuzzy set-based polynomial neurons (FSPNs)-based FSPNN with polynomial neurons (PNs)-based PNN. The design of the conventional genetically optimized HFPNN exploits the extended Group Method of Data Handling (GMDH) whose some essential parameters of the network being tuned with the use of genetic algorithms throughout the overall development process. Two general optimization mechanisms are explored. First, the structural optimization is realized via GAs while the ensuing detailed parametric optimization is carried out in the setting of a standard least square method-based learning. The performance of the gHFSPNN is quantified through extensive experimentation where we considered a number of modeling benchmarks (synthetic and experimental data already experimented with in fuzzy or neurofuzzy modeling).     

基于代数神经网络的置换多项式判定网络学习算法

从神经元的运算特性入手,对神经元的激发函数,网络结构,学习目标三方面进行了推广,设计出了一类用于有限域上置换多项式判定的多项式神经网络模型,它们是单输入单输出的3层神经网络。给出了两类置换多项式判定的离散网络模型学习算法,该算法简单可行,易于实现。     

IG-based genetically optimized fuzzy polynomial neural networks with fuzzy set-based polynomial neurons

In this study, we introduce and investigate a new topology of fuzzy-neural networks—fuzzy polynomial neural networks (FPNN) that is based on a genetically optimized multiplayer perceptron with fuzzy set-based polynomial neurons (FSPNs). We also develop a comprehensive design methodology involving mechanisms of genetic optimization and information granulation. In the sequel, the genetically optimized FPNN (gFPNN) is formed with the use of fuzzy set-based polynomial neurons (FSPNs) composed of fuzzy set-based rules through the process of information granulation. This granulation is realized with the aid of the C-means clustering (C-Means). The design procedure applied in the construction of each layer of an FPNN deals with its structural optimization involving the selection of the most suitable nodes (or FSPNs) with specific local characteristics (such as the number of input variable, the order of the polynomial, the number of membership functions, and a collection of specific subset of input variables) and address main aspects of parametric optimization. Along this line, two general optimization mechanisms are explored. The structural optimization is realized via genetic algorithms (GAs) and HCM method whereas in case of the parametric optimization we proceed with a standard least square estimation (learning). Through the consecutive process of structural and parametric optimization, a flexible neural network is generated in a dynamic fashion. The performance of the designed networks is quantified through experimentation where we use two modeling benchmarks already commonly utilized within the area of fuzzy or neurofuzzy modeling.     

Global output feedback control for uncertain nonlinear feedforward systems

This paper considers a class of uncertain nonlinear feedforward systems with unknown constant growth rate, output polynomial function growth rate and system input function growth rate. Under the most general growth rate condition, only one dynamic gain is used to compensate simultaneously these three types of growth rates, an output feedback controller is constructed to guarantee the boundedness of closed-loop system states and the convergence of original system states.     

基于注意力机制和混合网络的小群体情绪识别

针对自然状态下小群体图像的情绪分类,提出基于面部、场景和骨架3种视觉线索的混合深度网络,分别利用3类卷积神经网络(convolutional neural networks,CNN)分支独立学习,通过决策融合获得最终的情绪分类。其中面部CNN通过注意力机制学习不同人脸的权重,获得整张图片关于人脸的特征表示,利用large-margin softmax (L-softmax)损失函数进行判别性学习;使用先进的姿势估计方法 OpenPose获得图像中所有人体骨架,作为基于骨架卷积神经网络的输入。考虑图片的场景信息,将整张图片作为基于场景CNN的输入。实验结果表明,改进模型对自然状态下3种类型的小群体情绪识别鲁棒,取得了较高的准确率。     

Evolutionary design of hybrid self-organizing fuzzy polynomial neural networks with the aid of information granulation

We introduce a new architecture of information granulation-based and genetically optimized Hybrid Self-Organizing Fuzzy Polynomial Neural Networks (HSOFPNN). Such networks are based on genetically optimized multi-layer perceptrons. We develop their comprehensive design methodology involving mechanisms of genetic optimization and information granulation. The architecture of the resulting HSOFPNN combines fuzzy polynomial neurons (FPNs) that are located at the first layer of the network with polynomial neurons (PNs) forming the remaining layers of the network. The augmented version of the HSOFPNN, “IG_gHSOFPNN”, for brief, embraces the concept of information granulation and subsequently exhibits higher level of flexibility and leads to simpler architectures and rapid convergence speed to optimal structure in comparison with the HSOFPNNs and SOFPNNs.

The GA-based design procedure being applied at each layer of HSOFPNN leads to the selection of preferred nodes of the network (FPNs or PNs) whose local characteristics (such as the number of input variables, the order of the polynomial, a collection of the specific subset of input variables, the number of membership functions for each input variable, and the type of membership function) can be easily adjusted. In the sequel, two general optimization mechanisms are explored. The structural optimization is realized via GAs whereas the ensuing detailed parametric optimization is afterwards carried out in the setting of a standard least square method-based learning. The obtained results demonstrate a superiority of the proposed networks over the existing fuzzy and neural models.     

基于PSO与LS-SVM的作物需水量预测

为了提高作物需水量预测精度,提出基于粒子群优化算法(PSO)优化最小二乘支持向量机(LS-SVM)的预测模型。该模型以空气湿度、温度、太阳辐射以及风速为输入,利用多项式核函数和径向基核函数的非负线性组合构造核函数,将粒子群优化算法(PSO)与交叉验证方法用于确定模型参数。实验结果表明与神经网络和随机森林相比,PSO优化的LS-SVM可获得更好的预测精度和泛化能力,可用于节水灌溉,具有较高的应用价值。     

Adaptive Fuzzy Neural Networks as identifiers of discrete-time nonlinear dynamic systems

An adaptive supervised learning scheme is proposed in this paper for training Fuzzy Neural Networks (FNN) to identify discrete-time nonlinear dynamical systems. The FNN constructs are neural-network-based connectionist models consisting of several layers that are used to implement the functions of a fuzzy logic system. The fuzzy rule base considered here consists of Takagi-Sugeno IF-THEN rules, where the rule outputs are realized as linear polynomials of the input components. The FNN connectionist model is functionally partitioned into three separate parts, namely, the premise part, which provides the truth values of the rule preconditional statements, the consequent part providing the rule outputs, and the defuzzification part computing the final output of the FNN construct. The proposed learning scheme is a two-stage training algorithm that performs both structure and parameter learning, simultaneously. First, the structure learning task determines the proper fuzzy input partitions and the respective precondition matching, and is carried out by means of the rule base adaptation mechanism. The rule base adaptation mechanism is a self-organizing procedure which progressively generates the proper fuzzy rule base, during training, according to the operating conditions. Having completed the structure learning stage, the parameter learning is applied using the back-propagation algorithm, with the objective to adjust the premise/consequent parameters of the FNN so that the desired input/output representation is captured to an acceptable degree of accuracy. The structure/parameter training algorithm exhibits good learning and generalization capabilities as demonstrated via a series of simulation studies. Comparisons with conventional multilayer neural networks indicate the effectiveness of the proposed scheme.     

Self-organizing neurofuzzy networks based on evolutionary fuzzy granulation

Experimental software datasets describing software projects in terms of their complexity and development time have been a subject of intensive modeling. A number of various modeling methodologies and modeling designs have been proposed including such development frameworks as neural networks, fuzzy and neurofuzzy models. In this study, we introduce a concept of self-organizing neurofuzzy networks (SONFN), a hybrid modeling architecture combining neurofuzzy networks (NFN) and polynomial neural networks (PNN). For these networks we develop a comprehensive design methodology. The construction of SONFNs takes advantage of the well-established technologies of computational intelligence (CI), namely fuzzy sets, neural networks and genetic algorithms. The architecture of the SONFN results from a synergistic usage of NFNs and PNNs. NFN contributes to the formation of the premise part of the rule-based structure of the SONFN. The consequence part of the SONFN is designed using PNNs. We discuss two types of SONFN architectures whose taxonomy is based on the NFN scheme being applied to the premise part of SONFN. We introduce a comprehensive learning algorithm. It is shown that this network exhibits a dynamic structure as the number of its layers as well as the number of nodes in each layer of the SONFN are not predetermined (as this is the case in a popular topology of a multilayer perceptron). The experimental results include a well-known NASA dataset concerning software cost estimation.     

Continuous-time Wiener system identification

A continuous-time Wiener system is identified. The system consists of a linear dynamic subsystem and a memoryless nonlinear one connected in a cascade. The input signal is a stationary white Gaussian random process. The system is disturbed by stationary white random Gaussian noise. Both subsystems are identified from input-output observations taken at the input and output of the whole system. The a priori information is very small and, therefore, resulting identification problems are nonparametric. The impulse impulse of the linear part is recovered by a correlation method, while the nonlinear characteristic is estimated with the help of the nonparametric kernel regression method. The authors prove convergence of the proposed identification algorithms and examine their convergence rates     

Approximation of algebraic and trigonometric polynomials by feedforward neural networks

This paper presents a function approximation to a general class of polynomials by using one-hidden-layer feedforward neural networks(FNNs). Both the approximations of algebraic polynomial and trigonometric polynomial functions are discussed in details. For algebraic polynomial functions, an one-hidden-layer FNN with chosen number of hidden-layer nodes and corresponding weights is established by a constructive method to approximate the polynomials to a remarkable high degree of accuracy. For trigonometric functions, an upper bound of approximation is therefore derived by the constructive FNNs. In addition, algorithmic examples are also included to confirm the accuracy performance of the constructive FNNs method. The results show that it improves efficiently the approximations of both algebraic polynomials and trigonometric polynomials. Consequently, the work is really of both theoretical and practical significance in constructing a one-hidden-layer FNNs for approximating the class of polynomials. The work also paves potentially the way for extending the neural networks to approximate a general class of complicated functions both in theory and practice.     

Graph partitioning using annealed neural networks

A new algorithm, mean field annealing (MFA), is applied to the graph-partitioning problem. The MFA algorithm combines characteristics of the simulated-annealing algorithm and the Hopfield neural network. MFA exhibits the rapid convergence of the neural network while preserving the solution quality afforded by simulated annealing (SA). The rate of convergence of MFA on graph bipartitioning problems is 10-100 times that of SA, with nearly equal quality of solutions. A new modification to mean-field annealing is also presented which supports partitioning graphs into three or more bins, a problem which has previously shown resistance to solution by neural networks. The temperature-behavior of MFA during graph partitioning is analyzed approximately and shown to possess a critical temperature at which most of the optimization occurs. This temperature is analogous to the gain of the neurons in a neural network and can be used to tune such networks for better performance. The value of the repulsion penalty needed to force MFA (or a neural network) to divide a graph into equal-sized pieces is also estimated.