Search

The article discusses the use of model predictive control (MPC) for drone flight control in unknown dynamic environments. MPC is a practical method for drone flight control due to its robustness against modeling errors and uncertainties. However, it can degrade in performance when faced with unknown environmental dynamics. The paper introduces a parameterized MPC approach called hyMPC that adapts to uncertain environmental conditions. It uses a novel policy search framework to train a high-level Gaussian policy and neural network policies to provide real-time decision variables. The effectiveness of hyMPC is validated through numerical simulations, achieving a 100% success rate in drone flight tests. The text discusses a framework for maneuvering drones through swinging gates using reinforcement learning (RL) and model predictive control (MPC). The framework utilizes episode-based policy search methods to acquire parameterized motor primitives and skills for controlling the drone. The RL-driven MPC framework incorporates a multi-layer perceptron to predict the gate's future state and a hybrid MPC cost function for trajectory optimization. The policy parameters are updated using a reward function that evaluates the quality of sampled trajectories. The proposed framework is validated through numerical simulations. This document presents a novel framework for navigating drones through swinging gates with unknown dynamics. The framework combines real-time Model Predictive Control (MPC) and offline Reinforcement Learning (RL) techniques. The approach involves training a neural network to predict gate motions and using an MLP to guide trajectory planning. The framework divides the traversal task into gate-following and gate-traversing [Extracted from the article]

Shows how seemingly minor differences in training and evaluation procedures used in some studies of recurrent neural networks as context free language predictors can lead to significant differences in apparent network performance. Experimental methods; Results of the study; Conclusion.

In order to satisfy the current market demands for shorter lead-time and high quality products manufacturing enterprises have to integrate their design, production and quality assurance functions. In a computerized environment these functions are manifested in CAD, CAM and AVI respectively. In the recent past, much success has been achieved in integrating CAD and CAM. It is widely accepted that feature recognition systems, both 2D and 3D, have been one of the main contributors to CADCAM integration. However the existing feature recognition systems could not link CADCAM and AVI and hence could not close the manufacturing loop that ensures the production of designer intended features. This is mainly because the input formats of existing feature recognition systems and AVI systems have been different in format and structure. The system described in this paper attempts to redress this deficit at least partially by using a monochrome bitmap as a generic input, to which both CAD models and vision images could be converted. Hence the input to the system takes a form of third angle orthographic views, however, without hidden lines in order to facilitate dealing with vision images. This, in turn, augments difficulty faced in recognizing features. To overcome this difficulty, the help of evolutionary computing and artificial intelligence is sought in this system. This paper outlines the basis of a 2D feature recognition system that uses artificial neural networks along with a chain code method for eliciting feature information from monochrome bitmap of either vision images or CAD inputs hence providing a generic framework to integrate CADCAM and AVI. With further improvement to deal with geometry of the features, annotations and symbols this system could also help to salvage the massive store of engineering knowledge that exists in 2D form. [ABSTRACT FROM AUTHOR]

We present a comprehensive review of the evolutionary design of neural network architectures. This work is motivated by the fact that the success of an Artificial Neural Network (ANN) highly depends on its architecture and among many approaches Evolutionary Computation, which is a set of global-search methods inspired by biological evolution has been proved to be an efficient approach for optimizing neural network structures. Initial attempts for automating architecture design by applying evolutionary approaches start in the late 1980s and have attracted significant interest until today. In this context, we examined the historical progress and analyzed all relevant scientific papers with a special emphasis on how evolutionary computation techniques were adopted and various encoding strategies proposed. We summarized key aspects of methodology, discussed common challenges, and investigated the works in chronological order by dividing the entire timeframe into three periods. The first period covers early works focusing on the optimization of simple ANN architectures with a variety of solutions proposed on chromosome representation. In the second period, the rise of more powerful methods and hybrid approaches were surveyed. In parallel with the recent advances, the last period covers the Deep Learning Era, in which research direction is shifted towards configuring advanced models of deep neural networks. Finally, we propose open problems for future research in the field of neural architecture search and provide insights for fully automated machine learning. Our aim is to provide a complete reference of works in this subject and guide researchers towards promising directions. [ABSTRACT FROM AUTHOR]

The human arm exhibits outstanding manipulability in executing various tasks by taking advantage of its intrinsic compliance, force sensation, and tactile contact clues. By examining human strategy in controlling arm impedance, we may be able to understand underlying human motor control and develop control methods for dexterous robotic manipulation. This paper presents a novel method for estimating multijoint stiffness by using electromyogram (EMG) and an artificial neural network model. The artificial network model developed in this paper relates EMG data and joint motion data to joint stiffness. With the proposed method, the multijoint stiffness of the arm was estimated without complex calculation or specialized apparatus. The feasibility of the proposed method was confirmed through experimental and simulation results. [ABSTRACT FROM AUTHOR]

In recent years, interest in studying evolutionary algorithms (EAs) for dynamic optimization problems (DOPs) has grown due to its importance in real-world applications. Several approaches, such as the memory and multiple population schemes, have been developed for EAs to address dynamic problems. This paper investigates the application of the memory scheme for population-based incremental learning (PBIL) algorithms, a class of EAs, for DOPs. A PBIL-specific associative memory scheme, which stores best solutions as well as corresponding environmental information in the memory, is investigated to improve its adaptability in dynamic environments. In this paper, the interactions between the memory scheme and random immigrants, multipopulation, and restart schemes for PBILs in dynamic environments are investigated. In order to better test the performance of memory schemes for PBILs and other EAs in dynamic environments, this paper also proposes a dynamic environment generator that can systematically generate dynamic environments of different difficulty with respect to memory schemes. Using this generator, a series of dynamic environments are generated and experiments are carried out to compare the performance of investigated algorithms. The experimental results show that the proposed memory scheme is efficient for PBILs in dynamic environments and also indicate that different interactions exist between the memory scheme and random immigrants, multipopulation schemes for PBILs in different dynamic environments. [ABSTRACT FROM AUTHOR]

In this paper, stability and bifurcation of a general recurrent neural network with multiple time delays is considered, where all the variables of the network can be regarded as bifurcation parameters. It is found that Hopf bifurcation occurs when these parameters pass through some critical values where the conditions for local asymptotical stability of the equilibrium are not satisfied. By analyzing the characteristic equation and using the frequency domain method, the existence of Hopf bifurcation is proved. The stability of bifurcating periodic solutions is determined by the harmonic balance approach, Nyquist criterion, and graphic Hopf bifurcation theorem. Moreover, a critical condition is derived under which the stability is not guaranteed, thus a necessary and sufficient condition for ensuring the local asymptotical stability is well understood, and from which the essential dynamics of the delayed neural network are revealed. Finally, numerical results are given to verify the theoretical analysis, and some interesting phenomena are observed and reported. [ABSTRACT FROM AUTHOR]

The development of wind power generation has rapidly progressed over the last decade. With the advancement in wind turbine technology, wind energy has become competitive with other fuel-based resources. The fluctuation of wind, however, makes it difficult to optimize the usage of wind power. The current practice ignores wind generation capacity in the unit commitment (UC), which discounts its usable capacity and may cause operational issues when the installation of wind generation equipment increases. To ensure system reliability, the forecasting uncertainty must be considered in the incorporation of wind power capacity into generation planning. This paper discusses the development of an artificial-neural-network-based wind power forecaster and the integration of wind forecast results into UC scheduling considering forecasting uncertainty by the probabilistic concept of confidence interval. The data from a wind farm located in Lawton City, OK, is used in this paper. [ABSTRACT FROM AUTHOR]

A long-term goal of AI is to produce agents that can learn a diversity of skills throughout their lifetimes and continuously improve those skills via experience. A longstanding obstacle towards that goal is catastrophic forgetting, which is when learning new information erases previously learned information. Catastrophic forgetting occurs in artificial neural networks (ANNs), which have fueled most recent advances in AI. A recent paper proposed that catastrophic forgetting in ANNs can be reduced by promoting modularity, which can limit forgetting by isolating task information to specific clusters of nodes and connections (functional modules). While the prior work did show that modular ANNs suffered less from catastrophic forgetting, it was not able to produce ANNs that possessed task-specific functional modules, thereby leaving the main theory regarding modularity and forgetting untested. We introduce diffusion-based neuromodulation, which simulates the release of diffusing, neuromodulatory chemicals within an ANN that can modulate (i.e. up or down regulate) learning in a spatial region. On the simple diagnostic problem from the prior work, diffusion-based neuromodulation 1) induces task-specific learning in groups of nodes and connections (task-specific localized learning), which 2) produces functional modules for each subtask, and 3) yields higher performance by eliminating catastrophic forgetting. Overall, our results suggest that diffusion-based neuromodulation promotes task-specific localized learning and functional modularity, which can help solve the challenging, but important problem of catastrophic forgetting. [ABSTRACT FROM AUTHOR]

Artificial intelligence and all its supporting tools, e.g. machine and deep learning in computational intelligence-based systems, are rebuilding our society (economy, education, life-style, etc.) and promising a new era for the social welfare state. In this paper we summarize recent advances in data science and artificial intelligence within the interplay between natural and artificial computation. A review of recent works published in the latter field and the state the art are summarized in a comprehensive and self-contained way to provide a baseline framework for the international community in artificial intelligence. Moreover, this paper aims to provide a complete analysis and some relevant discussions of the current trends and insights within several theoretical and application fields covered in the essay, from theoretical models in artificial intelligence and machine learning to the most prospective applications in robotics, neuroscience, brain computer interfaces, medicine and society, in general. [ABSTRACT FROM AUTHOR]

In the literature, it was reported that the chaotic artificial neural network model with sinusoidal activation functions possesses a large memory capacity as well as a remarkable ability of retrieving the stored patterns, better than the conventional chaotic model with only monotonic activation functions such as sigmoidal functions. This paper, from the viewpoint of the anti-integrable limit, elucidates the mechanism inducing the superiority of the model with periodic activation functions that includes sinusoidal functions. Particularly, by virtue of the anti-integrable limit technique, this paper shows that any finite-dimensional neural network model with periodic activation functions and properly selected parameters has much more abundant chaotic dynamics that truly determine the model's memory capacity and pattern-retrieval ability. To some extent, this paper mathematically and numerically demonstrates that an appropriate choice of the activation functions and control scheme can lead to a large memory capacity and better pattern-retrieval ability of the artificial neural network models. [ABSTRACT FROM AUTHOR]

This paper is concerned with the stability analysis of discrete-time recurrent neural networks (RNNS) with time delays as random variables drawn from some probability distribution. By introducing the variation probability of the time delay, a common delayed discrete-time RNN system is transformed into one with stochastic parameters. Improved conditions for the mean square stability of these systems are obtained by employing new Lyapunov functions and novel techniques are used to achieve delay dependence. The merit of the proposed conditions lies in its reduced conservatism, which is made possible by considering not only the range of the time delays, but also the variation probability distribution. A numerical example is provided to show the advantages of the proposed conditions. Stability Analysis of Discrete-Time Recurrent Neural Networks With Stochastic Delay. [ABSTRACT FROM AUTHOR]

In this paper, a universal simulator for cellular neural network (CNN) is presented. This simulator is capable of performing Raster simulation for any size of input image, and thus is a powerful tool for researchers investigating potential applications of CNN. This paper reports the latency properties of CNNs along with popular numerical integration algorithms; results and comparisons are also presented. [ABSTRACT FROM AUTHOR]

This paper applies artificial neural networks (ANNs) trained with a multiobjective algorithm to preprocess the ground penetrating radar data obtained from a finite-difference time-domain (FDTD) model. This preprocessing aims at improving the target's reflected wave signal-to-noise ratio (SNR). Once trained, the NN behaves as an adaptive filter which minimizes the cross-validation error. Results considering both White and colored Gaussian noise, with many different SNR, are presented and they show the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

In this paper, regularization path algorithms were proposed as a novel approach to the model selection problem by exploring the path of possibly all solutions with respect to some regularization hyperparameter in an efficient way. This approach was later extended to a support vector regression (SVR) model called ∈-SVR. However, the method requires that the error parameter ∈ be set a priori. This is only possible if the desired accuracy of the approximation can be specified in advance. In this paper, we analyze the solution space for ∈-SVR and propose a new solution path algorithm, called ∈-path algorithm, which traces the solution path with respect to the hyperparameter ∈ rather than A. Although both two solution path algorithms possess the desirable piecewise linearity property, our ∈-path algorithm overcomes some limitations of the original λ-path algorithm and has more advantages. It is thus more appealing for practical use. [ABSTRACT FROM AUTHOR]

Abstract-The robust state estimation problem for a class of uncertain neural networks with time-varying delay is studied in this paper. The parameter uncertainties are assumed to be norm bounded. Based on a new bounding technique, a sufficient condition is presented to guarantee the existence of the desired state estimator for the uncertain delayed neural networks. The criterion is dependent on the size of the time-varying delay and on the size of the time derivative of the time-varying delay. It. is shown that the design of the robust state estimator for such neural networks can be achieved by solving a linear matrix inequality (LMI), which can be easily facilitated by using some standard numerical packages. Finally, two simulation examples are given to demonstrate the effectiveness of the developed approach. [ABSTRACT FROM AUTHOR]

In this paper, the recurrent neural networks (RNNs) with a generalized activation function class is proposed. In this proposed model, every component of the neuron's activation function belongs to a convex hull which is bounded by two odd symmetric piecewise linear functions that are convex or concave over the real space. All of the convex hulls are composed of generalized activation function classes. The novel activation function class is not only with a more flexible and more specific description of the activation functions than other function classes but it also generalizes some traditional activation function classes. The absolute exponential stability (AEST) of the RNN with a generalized activation function class is studied through three steps. The first step is to demonstrate the global exponential stability (GES) of the equilibrium point of original RNN with a generalized activation function being equivalent to that of RNN under all vertex functions of convex hull. The second step transforms the RNN under every vertex activation function into neural networks under an array of saturated linear activation functions. Because the GES of the equilibrium point of three systems are equivalent, the next stability analysis focuses on the GES of the equilibrium point of RNN system under an array of saturated linear activation functions. The last step is to study both the existence of equilibrium point and the GES of the RNN under saturated linear activation functions using the theory of M-matrix. In the end, a two-neuron RNN with a generalized activation function is constructed to show the effectiveness of our results. [ABSTRACT FROM AUTHOR]

An important consideration when applying neural networks is the sensitivity to weights and threshold in strict separating systems representing a linearly separable function. Perturbations may affect weights and threshold so that it is important to estimate the maximal percentage error in weights and threshold, which may be allowed without altering the linearly separable function. In this paper, we provide the greatest allowed bound which can be associated to every strict separating system representing a linearly separable function. The proposed bound improves the tolerance that Hu obtained. Furthermore, it is the greatest bound for any strict separating system. This is the reason why we call it the greatest tolerance. [ABSTRACT FROM AUTHOR]

In this paper, the robust synchronization problem is addressed for recurrent neural networks with time-varying delay by linear feedback control. Robustness in the present paper is referred to as the allowance of parameters mismatch between the drive system and the response system. Sufficient conditions for robust synchronization with a synchronization error bound, expressed as linear matrix inequality (LMI), are derived based on Lyapunov–Krasovskii functionals. Both delay-dependent and delay-independent conditions are obtained. Two examples are given to illustrate the results. [ABSTRACT FROM AUTHOR]

Principal component analysis (PCA) using neural networks is an active research field with many applications to signal processing and data analysis. This paper presents a PCA neural network endowed with a novel learning algorithm, and an analysis of its features. As the basic discrete-time Oja's PCA neural network does not converge globally, it is important to derive a robust and globally convergent PCA learning algorithm. Based on previous works on the globally convergent PCA learning algorithm, a robust and globally convergent PCA learning algorithm is proposed in this paper. The behavior of this discrete-time learning algorithm is directly studied in this paper. We show that the algorithm is robust and globally convergent, The selection of the parameters of this algorithm will also be discussed in details. Finally, simulation results are provided to verify the theoretical results presented. Compared to other PCA learning algorithms, the proposed algorithm performs favorably in terms of robust stability, global convergence, speed and accuracy. [ABSTRACT FROM AUTHOR]

This paper presents a cooperative coevolutive approach for designing neural network ensembles. Cooperative coevolution is a recent paradigm in evolutionary computation that allows the effective modeling of cooperative environments. Although theoretically, a single neural network with a sufficient number of neurons in the hidden layer would suffice to solve any problem, in practice many real-world problems are too hard to construct the appropriate network that solve them. In such problems, neural network ensembles are a successful alternative. Nevertheless, the design of neural network ensembles is a complex task. In this paper, we propose a general framework for designing neural network ensembles by means of cooperative coevolution. The proposed model has two main objectives: first, the improvement of the combination of the trained individual networks; second, the cooperative evolution of such networks, encouraging collaboration among them, instead of a separate training of each network. In order to favor the cooperation of the networks, each network is evaluated throughout the evolutionary process using a multiobjective method. For each network, different objectives are defined, considering not only its performance in the given problem, but also its cooperation with the rest of the networks. In addition, a population of ensembles is evolved, improving the combination of networks and obtaining subsets of networks to form ensembles that perform better than the combination of all the evolved networks. The proposed model is applied to ten real-world classification problems of a very different nature from the UCI machine learning repository and proben1 benchmark set. In all of them the performance of the model is better than the performance of standard ensembles in terms of generalization error. Moreover, the size of the obtained ensembles is also smaller. [ABSTRACT FROM AUTHOR]

Determination of feeder configuration involving the switch statuses is important for the operation in the distribution system. Minimization of the megawatt loss is crucial if the system is in a normal condition; the voltage drop is expected to be as low as possible while a fault occurs in the system. This paper presents a method based on genetic algorithms (GAs) to determine the network configuration. The proposed method, formulated as a fuzzy multiobjective problem, takes both the normal condition and the contingencies (faults) into account. The switch statuses are considered as preventive controls for multiple individual balanced faults. A vertex encoding based on the Prufer number is used in GA for encoding the chromosomes (strings). The Prufer number ensures that the system structure will be radial for the distribution system. The simulation results from two distribution systems are used to show the applicability of the proposed method. [ABSTRACT FROM AUTHOR]

In this paper, we deal with encoding and enumerating threshold functions defined on n-dimensional binary inputs. The paper specifies situations in which the unique characterization of functions from a given class is preserved by usage of an appropriate set of discrete moments. Moreover, sometimes such a characterization (coding) is optimal with respect to the number of necessary bit rate per coded function. By estimating the number of possible values of the discrete moments used, several upper bounds (for different classes of threshold functions) are derived, some of which are better than those previously known. [ABSTRACT FROM AUTHOR]

A fast target maneuver detecting and highly accurate tracking technique using a neural fuzzy network based on Kalman filter is proposed in this paper. In the automatic target tracking system, there exists an important and difficult problem: how to detect the target maneuvers and fast response to avoid miss-tracking? The traditional maneuver detection algorithms, such as variable dimension filter (VDF) and input estimation (IE) etc., are computation intensive and difficult to implement in real time. To solve this problem, neural network algorithms have been issued recently. However, the normal neural networks such as backpropagation networks usually produce the extra problems of low convergence speed and/or large network size. Furthermore, the way to decide the network structure is heuristic. To overcome these defects and to make use of neural learning ability, a developed standard Kalman filter with a self-constructing neural fuzzy inference network (KF-SONFIN) algorithm for target tracking is presented in this paper. By generating possible target trajectories including maneuver information to train the SONFIN, the trained SONFIN can detect when the maneuver occurred, the magnitude of maneuver values and when the maneuver disappeared. Without having to change the structure of Kalman filter nor modeling the maneuvering target, this new algorithm, SONFIN, can always find itself an economic network size with a fast learning process. Simulation results show that the KF-SONFIN is superior to the traditional IE and VDF methods in estimation accuracy. [ABSTRACT FROM AUTHOR]

This paper presents the application of FeaSANNT, an evolutionary algorithm for optimization of artificial neural networks, to the training of a multi-layer perceptron for identification of defects in wood veneer. Given a fixed artificial neural network structure, FeaSANNT concurrently evolves the input feature vector and the network weights. The novelty of the method lies in the implementation of the embedded approach in an evolutionary feature selection paradigm. Experimental tests show that the proposed algorithm produces high-performing solutions with robust learning results. A significant reduction of the set of veneer features is obtained. Experimental comparisons are made with a previous method based on statistical filtering of the input features and a standard genetic wrapper algorithm. In the first case, FeaSANNT greatly reduces the feature set with no degradation of the neural network accuracy. Moreover, FeaSANNT entails lower design costs, since feature selection is fully automated. In the second case, the proposed algorithm achieves superior results in terms of identification accuracy and reduction of the feature set. FeaSANNT involves also lower computational costs than the standard evolutionary wrapper approach and eases the algorithm design effort. Limited overlapping is observed between the patterns of features selected by the three algorithms. This result suggests that the full feature set contains mainly redundant attributes. [ABSTRACT FROM AUTHOR]

Predicting corporate failure or bankruptcy is one of the most important problems facing business and government. The recent Savings and Loan crisis is one example, where bankruptcies cost the United States billions of dollars and became a national political issue. This paper provides a 'meta analysis' of the use of neural networks to predict corporate failure. Fifteen papers are reviewed and compared in order to investigate 'what works and what doesn't work'. The studies are compared for their formulations including aspects such as the impact of using different percentages of bankrupt firms, the software they used, the input variables, the nature of the hidden layer used, the number of nodes in the hidden layer, the output variables, training and testing and statistical analysis of results. Then the findings are compared across a number of dimensions, including, similarity of comparative solutions, number of correct classifications, impact of hidden layers, and the impact of the percentage of bankrupt firms. [ABSTRACT FROM AUTHOR]

A threshold gate is a theoretical model of a neuron, and a threshold circuit is a theoretical model of a neural network. The energy e of a threshold circuit C is defined to be the maximum number of gates outputting ones, where the maximum is taken over all input assignments to C. In this paper, we prove that the comparison function of 2n variables is computable by a polynomial-weight threshold circuit C of energy e, size s =O(n=log n), depth d = O(n/(e log n)) for any e, 3 ⩽ e ⩽ n=dlog ne. Our result implies that one can construct an energy-efficient circuit computing the comparison function if it is allowable to use large depth. [ABSTRACT FROM AUTHOR]

This article presents the issues of an examination and evaluation of the reliability of a technical object after its regeneration in a maintenance system. The preventive activities (regeneration) of an object are conducted in a maintenance system which includes an artificial neural network. The author made an attempt in the paper to prove that an artificial neural network has an indirect influence on the quality of preventive activities organized for a technical object and hence on the level of its reliability. This influence is the result of the fact that information developed by a neural network constitutes the basis in the process of the generation of an expert knowledge base on the basis of which the maintenance system is organized. This paper also covers the theoretical basis. Those quantities were defined which may be used in reliability tests of technical objects. The possibilities were presented of testing the reliability: the quality of the use of an object after its regeneration. These tests may be carried out in a short period of time and in a longer period of time after the preventive activities have been carried out for a device. This article also presents the idea of a reliability test conducted for a technical object on the example of a device which controls the work of a petrol engine. For this purpose, the structure of a maintenance system was designed on the basis of an expert knowledge base determined for an object. Diagnostic information from an artificial neural network constitutes the basis for these activities. A substantial part of the article is devoted to the description of the method of the reliability test of the object and to an analysis of the results obtained. The test was carried out in a simulation manner after a long period of time following regeneration of a technical object. [ABSTRACT FROM AUTHOR]

The detection of the malfunction of an individual robot is an essential issue in the control of multirobot systems. In this paper, a malfunction is defined in terms of a robot’s motion performance. A monitoring system to detect robot malfunctions is developed in the form of a centralized high-level planner. The monitoring system collects data on the positions and any formation errors of each robot and evaluates working performance in terms of current and past states. A back-propagation neural-network-based controller is developed to evaluate the status of the robots, i.e. working or failed. The neural network is trained using both positive and negative examples. Simulations and experiments are performed to demonstrate the effectiveness of the proposed approach. [ABSTRACT FROM AUTHOR]

This paper presents hardware implementation of Artificial Neural Networks (ANN) that are used for human hand's natural gestures recognition. Main goal of this project is to implement a recognition system that recognizes data gathered from various sensors placed on a bracelet. This easy to use interface can be used even by elderly or impaired people to control electronic/electric devices only with hand gestures. [ABSTRACT FROM AUTHOR]

The existing Multi-View Learning (MVL) is to discuss how to learn from patterns with multiple information sources and has been proven its superior generalization to the usual Single-View Learning (SVL). However, in most real-world cases there are just single source patterns available such that the existing MVL cannot work. The purpose of this paper is to develop a new multi-view regularization learning for single source patterns. Concretely, for the given single source patterns, we first map them into M feature spaces by M different empirical kernels, then associate each generated feature space with our previous proposed Discriminative Regularization (DR), and finally synthesize M DRs into one single learning process so as to get a new Multi-view Discriminative Regularization (MVDR), where each DR can be taken as one view of the proposed MVDR. The proposed method achieves: (1) the complementarity for multiple views generated from single source patterns; (2) an analytic solution for classification; (3) a direct optimization formulation for multi-class problems without one-against-all or one-against-one strategies. [ABSTRACT FROM AUTHOR]

The task scheduling problem in heterogeneous systems (TSPHS) is a NP-complete problem. It is a multiobjective optimization problem (MOP). The objectives such as makespan, average flow time, robustness and reliability of the schedule are considered for solving task scheduling problem. This paper considers only the two objectives of minimizing the makespan (schedule length) and maximizing the reliability in the multiobjective task scheduling problem. Multiobjective Evolutionary Computation algorithms (MOEAs) are well suited for Multiobjective task scheduling for heterogeneous environment. The two Multi-Objective Evolutionary Algorithms such as Multiobjective Genetic Algorithm (MOGA) and Multiobjective Evolutionary Programming (MOEP) with non-dominated sorting are developed and compared for the various random task graphs and also for a real-time numerical application graph. This paper demonstrates the capabilities of MOEAs to generate well-distributed Pareto optimal fronts in a single run. [ABSTRACT FROM AUTHOR]

Metaplasticity property of biological synapses is interpreted in this paper as the concept of placing greater emphasis on training patterns that are less frequent. A novel implementation is proposed in which, during the network learning phase, a priority is given to weight updating of less frequent activations over the more frequent ones. Modeling this interpretation in the training phase, the hypothesis of an improved training is tested on the Multilayer Perceptron type network with Backpropagation training. The results obtained for the chosen application show a much more efficient training, while at least maintaining the Multilayer Perceptron performance. [ABSTRACT FROM AUTHOR]

In the paper, the anti-synchronization problem of the general delayed chaotic neural networks is investigated. For the master and slave systems, we obtain a control law to achieve the state anti-synchronization of two identical chaotic neural networks. By using the Halanay inequality lemma and Lyapunov stability method, we derive a delay independent sufficient exponential anti-synchronization condition relative to the parameters of the systems and controller gain matrix. The condition is easily verified in practice. Finally, the theoretical results are applied to two delayed chaotic neural networks, and numerical simulations are given to demonstrate the performance of the proposed scheme throughout some examples. [ABSTRACT FROM AUTHOR]

This paper uses the data from seven wind farms at Muppandal, Tamil Nadu, India, collected for three years from April 2002 to March 2005 for the estimation of energy yield from wind farms. The model is developed with the help of neural network methodology, and it involves three input variables-wind speed, relative humidity, and generation hours-and one output variable, which give the energy output from wind farms. The modeling is done using MATLAB software. The most appropriate neural network configuration after trial and error is found to be 3-5-1 (3 input layer neurons, 5 hidden layer neurons, 1 output layer neuron). The mean square error for the estimated values with respect to the measured data is 7.6 × 10-3. The results demonstrate that this work is an efficient energy yield estimation tool for wind farms. [ABSTRACT FROM AUTHOR]

This paper presents a coupled approach using an artificial neural network (ANN) and the finite difference method (FDM) that has been developed to predict the distribution of axial load along fully grouted standard cable bolts in the field using laboratory pullout test data. A back-propagation training algorithm was used in ANN to determine axial loads in the cables tested in the laboratory. The ANN component of the computational model was trained using two different types of data sets. At first, the ANN was trained to predict the axial loads in a series of short cables grouted with Portland cement at a specific water-to-cement ratio and subjected to different radial confining stiffness values. Next, the ANN model was trained for an expanded case to include the influence of lateral confining stress on the distribution of axial load in the cable reinforcement. Finally, the ANN model was implemented into a widely used, FDM-based geotechnical software (FLAC). The accuracy of the ANN–FDM model is demonstrated in this paper against measured data from laboratory and field tests. The analysis approach introduced in this study is a valuable computational tool that can be used to determine the axial load distribution in long standard cable bolts, which are commonly installed to stabilize rock masses in various geotechnical, transportation, and mining applications. [ABSTRACT FROM AUTHOR]

Abstract  Metaheuristics are general algorithmic frameworks, often nature-inspired, designed to solve complex optimization problems, and they are a growing research area since a few decades. In recent years, metaheuristics are emerging as successful alternatives to more classical approaches also for solving optimization problems that include in their mathematical formulation uncertain, stochastic, and dynamic information. In this paper metaheuristics such as Ant Colony Optimization, Evolutionary Computation, Simulated Annealing, Tabu Search and others are introduced, and their applications to the class of Stochastic Combinatorial Optimization Problems (SCOPs) is thoroughly reviewed. Issues common to all metaheuristics, open problems, and possible directions of research are proposed and discussed. In this survey, the reader familiar to metaheuristics finds also pointers to classical algorithmic approaches to optimization under uncertainty, and useful informations to start working on this problem domain, while the reader new to metaheuristics should find a good tutorial in those metaheuristics that are currently being applied to optimization under uncertainty, and motivations for interest in this field. [ABSTRACT FROM AUTHOR]

Abstract  A novel scheme of digital image watermarking based on the combination of dual-tree wavelet transform (DTCWT) and probabilistic neural network is proposed in this paper. Firstly, the original image is decomposed by DTCWT, and then the watermark bits are added to the selected coefficients blocks. Because of the learning and adaptive capabilities of neural networks, the trained neural networks can recover the watermark from the watermarked images. Experimental results show that the proposed scheme has good performance against several attacks. [ABSTRACT FROM AUTHOR]

This paper develops and presents a new neural network approach to model trip distribution, which is one of the important phases of conventional four-step travel demand modelling. The trip distribution problem has been investigated using back-propagation artificial neural networks in a number of studies and it was concluded that back-propagation artificial neural networks underperform when compared to traditional models. Such underperformance is due to the thresholding of the linearly combined inputs by utilising a non-linear function and carrying out this operation both in hidden and output layers. The proposed neural trip distribution model does not threshold the linearly combined outputs from the hidden layer. This makes it different from back-propagation artificial neural networks where combined inputs from the hidden layer are activated once more in the output layer. In addition, the neuron in the output layer is used as a summation unit in contrast to the methodologies cited in the neural network applications literature. At the same time, the bias neuron is not connected to the output neuron in the output layer. When this model is compared with various approaches such as the gravity model, modular neural networks and back-propagation neural networks, it was concluded that this new model provides better prediction of trip distribution and therefore, outperforms all the existing approaches. [ABSTRACT FROM AUTHOR]

The increasing popularity of e-learning has created a need for accurate student achievement prediction mechanisms, allowing instructors to improve the efficiency of their courses by addressing specific needs of their students at an early stage. In this paper, a student achievement prediction method applied to a 10-week introductory level e-learning course is presented. The proposed method uses multiple feed-forward neural networks to dynamically predict students' final achievement and to cluster them in two virtual groups, according to their performance. Multiple-choice test grades were used as the input data set of the networks. This form of test was preferred for its objectivity. Results showed that accurate prediction is possible at an early stage, more specifically at the third week of the 10-week course. In addition, when students were clustered, low misplacement rates demonstrated the adequacy of the approach. The results of the proposed method were compared against those of linear regression and the neural-network approach was found to be more effective in all prediction stages. The proposed methodology is expected to support instructors in providing better educational services as well as customized assistance according to students' predicted level of performance. [ABSTRACT FROM AUTHOR]

This paper presents a new method of online estimation of the stator and rotor resistance of the induction motor in the indirect vector-controlled drive, with artificial neural networks. The back propagation algorithm is used for training of the neural networks. The error between the rotor flux linkages based on a neural network model and a voltage model is back propagated to adjust the weights of the neural network model for the rotor resistance estimation. For the stator resistance estimation, the error between the measured stator current and the estimated stator current using neural network is back propagated to adjust the weights of the neural network. The performance of the stator and rotor resistance estimators and torque and flux responses of the drive, together with these estimators, is investigated with the help of simulations for variations in the stator and rotor resistance from their nominal values. Both types of resistance are estimated experimentally, using the proposed neural network in a vector-controlled induction motor drive. Data on tracking performances of these estimators are presented. With this approach, the rotor resistance estimation was found to be insensitive to the stator resistance variations both in simulation and experiment. [ABSTRACT FROM AUTHOR]

In this paper, a new kernel-based method for data visualization and dimensionality reduction is proposed. A reference point is considered corresponding to additional constraints taken in the problem formulation. In contrast with the class of kernel eigenmap methods, the solution (coordinates in the low-dimensional space) is characterized by a linear system instead of an eigenvalue problem. The kernel maps with a reference point are generated from a least squares support vector machine (LS-SVM) core part that is extended with an additional regularization term for preserving local mutual distances together with reference point constraints. The kernel maps possess primal and dual model representations and provide out-of-sample extensions, e.g., for validation-based tuning. The method is illustrated on toy problems and real-life data sets. [ABSTRACT FROM AUTHOR]

Evolutionary algorithms are a class of stochastic search methods that attempts to emulate the biological process of evolution, incorporating concepts of selection, reproduction, and mutation. In recent years, there has been an increase in the use of evolutionary approaches in the training of artificial neural networks (ANNs). While evolutionary techniques for neural networks have shown to provide superior performance over conventional training approaches, the simultaneous optimization of network performance and architecture will almost always result in a slow training process due to the added algorithmic complexity. In this paper, we present a geometrical measure based on the singular value decomposition (SVD) to estimate the necessary number of neurons to be used in training a single-hidden-layer feedforward neural network (SLFN). In addition, we develop a new hybrid multiobjective evolutionary approach that includes the features of a variable length representation that allow for easy adaptation of neural networks structures, an architectural recombination procedure based on the geometrical measure that adapts the number of necessary hidden neurons and facilitates the exchange of neuronal information between candidate designs, and a microhybrid genetic algorithm (μHGA) with an adaptive local search intensity scheme for local fine-tuning. In addition, the performances of well-known algorithms as well as the effectiveness and contributions of the proposed approach are analyzed and validated through a variety of data set types. [ABSTRACT FROM AUTHOR]

The FastICA algorithm is one of the most prominent methods to solve the problem of linear independent component analysis (ICA). Although there have been several attempts to prove local convergence properties of FastICA, rigorous analysis is still missing in the community. The major difficulty of analysis is because of the well-known sign-flipping phenomenon of FastICA, which causes the discontinuity of the corresponding FastICA map on the unit sphere. In this paper, by using the concept of principal fiber bundles, FastICA is proven to be locally quadratically convergent to a correct separation. Higher order local convergence properties of FastICA are also investigated in the framework of a scalar shift strategy. Moreover, as a parallelized version of FastICA, the so-called QR FastICA algorithm, which employs the QR decomposition (Gram-Schmidt orthonormalization process) instead of the polar decomposition, is shown to share similar local convergence properties with the original FastICA. [ABSTRACT FROM AUTHOR]