The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
In this paper, a concept of balanced learning is presented, and an improved neural networks learning scheme is proposed to speed up the learning process in cerebellar model articulation controllers (CMAC). In the conventional CMAC learning scheme, the corrected amounts of errors are equally distributed into all addressed hypercubes, regardless of the credibility of those hypercubes. The proposed improved...
A novel cooperative evolutionary system, i.e., CGPNN, for automatic design artificial neural networks (ANN’s) is presented where ANN’s structure and parameters are tuned simultaneously. The algorithms used in CGPNN combine genetic algorithm (GA) and particle swarm optimization (PSO) on the basis of a direct encoding scheme. In CGPNN, standard (real-coded) PSO is employed to training ANN’s free parameters...
Medical Image Analysis represents a very important step in clinical diagnosis. It provides image segmentation of the Region of Interest (ROI) and the generation of a three-dimensional model, representing the selected object. In this work, was proposed a neural network segmentation based on Self-Organizing Maps (SOM) and a three-dimensional SOM architecture to create a 3D model, starting from 2D data...
In this paper, a new neural network for a class of variational inequalities with linear and nonlinear constraints is proposed by converting it into an extended variational inequality. The proposed neural network with the asymmetric mapping is proved to be stable in the sense of Lyapunov and converge to a solution of the original problem within a finite time under a weaker co-coercivity condition by...
Support vector machine (SVM), proposed by Vapnik based on statistical learning theory, is a novel machine learning method. However, there are two problems to be solved in this field: one is the multi-class classification problem, and the other is the sensitivity to the noisy data. In order to overcome these difficulties, a novel method of fuzzy compensation multi-class support vector machine, named...
In this paper, three existing chaotic neural network models are reviewed and the searching ability of these models is analyzed, a novel chaotic neural network with varying initial value is proposed to solve problems of the lower convergence rate and long searching time in the existing method. It is different from the other modified chaotic neural networks in the aspect that it seeks the better initial...
The chaotic synchronizations of Hindmarsh-Rose (HR) neurons networks linked using the nonlinear coupling feedback functions constructed specially are discussed. The method is an expansion of SC method based on the stability criterion. The stable chaotic synchronization can be achieved without calculation of the maximum Lyapunov exponent when the coupling strength is taken as reference value. The efficiency...
The paper proposes an edge-effect training multi-class fuzzy support vector machine (EFSVM). It treats the training data points with different importance in the training process, and especially emphasizes primary contribution of these points distributed in edge area of data sets for classification, and then assigns them greater fuzzy membership degrees, thus assures that the nearer these points are...
The paper presents an input-expansion-based improved method for general regression neural network (GRNN) and BP network. Using second-order inner product function or Chebyshev polynomial function to expand input vector of original samples, which makes input vector mapped into a higher-dimension pattern space and thus leads to the samples data more easily separable. The classification results for both...
An online gradient method with momentum for feedforward neural network is considered. The learning rate is set to be a constant and the momentum coefficient an adaptive variable. Both the weak and strong convergence results are proved, as well as the convergence rates for the error function and for the weight.
We propose a simple yet computationally efficient construction algorithm for two-class kernel classifiers. In order to optimise classifier’s generalisation capability, an orthogonal forward selection procedure is used to select kernels one by one by minimising the leave-one-out (LOO) misclassification rate directly. It is shown that the computation of the LOO misclassification rate is very efficient...
Chaotic neural networks have been proved to be powerful tools for escaping from local minima. In this paper, we first retrospect Chen’s chaotic neural network and then propose a novel Gauss-Morlet-Sigmoid chaotic neural network model. Second, we make an analysis of the largest Lyapunov exponents of the neural units of Chen’s and the Gauss-Morlet-Sigmoid model. Third, 10-city traveling salesman problem...
In this paper we use a novel neural approach for face recognition with Hidden Markov Models. A method based on the extraction of 2D-DCT feature vectors is described, and the recognition results are compared with a new face recognition approach with Artificial Neural Networks (ANN). ANNs are used to compress a bitmap image in order to represent it with a number of coefficients that is smaller than...
The application of neural network ensemble (NNE) to economic forecasting can heighten the generalization ability of learning systems through training multiple neural networks and combining their results. An improved principal component analysis (IPCA) is developed to extract the principal component of the economic data under the prerequisite that the main information of original economic data is not...
In this paper we propose two versions of Stacked Generalization as the combination module of an ensemble of neural networks. The first version only uses the information provided by expert networks. The second one uses the information provided by experts and the input data of the pattern that is being classified. Finally, we have performed a comparison among 6 classical combination methods and the...
This paper presents a novel approach to macro-economy forecasting based on the Fuzzy Neural Networks. This method employs the expert opinions, statistical analysis and the Genetic Algorithm, to enhance the model of Fuzzy Neural Network. Our method combines the expert opinions and the results of statistical analysis to determine the input parameters of the prediction model, and adopts the Genetic Algorithm...
A new approach to the construction and optimisation of ‘eng-genes’ grey-box neural networks is investigated. A forward selection algorithm is used to optimise both the network weights and biases and the parameters of the system-derived activation functions. The algorithm is used for both conventional neural network and eng-genes modelling of a simulated Continuously Stirred Tank Reactor. The resulting...
In this paper, we present the experimental works performed to test and explore the performance of our proposed framework: meta-learning evolutionary artificial neural network by means of cellular automata (MLEANN-CA). This framework based on evolutionary computation with direct and indirect encoding methods (cellular automata) for automatic design of optimal artificial neural networks wherein the...
To overcome the deficiency of Support Vector Machine (SVM) for regression, dynamic ε-SVM method was proposed. To establish precise mathematical models, a new modeling method was introduced, combining self-organizing feature map (SOFM) with the dynamic ε-SVM. Firstly, SOFM was used as a clustering algorithm to partition the whole input space into several disjointed regions; then, the dynamic ε-SVM...
In this paper, we view equalization as a multi-class classification problem and use neural networks to detect binary signals in the presence of noise and interference. In particular, we compare the performance of a recently published training algorithm, a multi-gradient, with that of the conventional back-propagation. Then, we apply a feature extraction to obtain more efficient neural networks. Experiments...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.