Basis Function Network (basis + function_network)

Distribution by Scientific Domains

Kinds of Basis Function Network

  • radial basis function network


  • Selected Abstracts


    Time series forecasting by combining the radial basis function network and the self-organizing map

    HYDROLOGICAL PROCESSES, Issue 10 2005
    Gwo-Fong Lin
    Abstract Based on a combination of a radial basis function network (RBFN) and a self-organizing map (SOM), a time-series forecasting model is proposed. Traditionally, the positioning of the radial basis centres is a crucial problem for the RBFN. In the proposed model, an SOM is used to construct the two-dimensional feature map from which the number of clusters (i.e. the number of hidden units in the RBFN) can be figured out directly by eye, and then the radial basis centres can be determined easily. The proposed model is examined using simulated time series data. The results demonstrate that the proposed RBFN is more competent in modelling and forecasting time series than an autoregressive integrated moving average (ARIMA) model. Finally, the proposed model is applied to actual groundwater head data. It is found that the proposed model can forecast more precisely than the ARIMA model. For time series forecasting, the proposed model is recommended as an alternative to the existing method, because it has a simple structure and can produce reasonable forecasts. Copyright 2005 John Wiley & Sons, Ltd. [source]


    Modelling small-business credit scoring by using logistic regression, neural networks and decision trees

    INTELLIGENT SYSTEMS IN ACCOUNTING, FINANCE & MANAGEMENT, Issue 3 2005
    Mirta Bensic
    Previous research on credit scoring that used statistical and intelligent methods was mostly focused on commercial and consumer lending. The main purpose of this paper is to extract important features for credit scoring in small-business lending on a dataset with specific transitional economic conditions using a relatively small dataset. To do this, we compare the accuracy of the best models extracted by different methodologies, such as logistic regression, neural networks (NNs), and CART decision trees. Four different NN algorithms are tested, including backpropagation, radial basis function network, probabilistic and learning vector quantization, by using the forward nonlinear variable selection strategy. Although the test of differences in proportion and McNemar's test do not show a statistically significant difference in the models tested, the probabilistic NN model produces the highest hit rate and the lowest type I error. According to the measures of association, the best NN model also shows the highest degree of association with the data, and it yields the lowest total relative cost of misclassification for all scenarios examined. The best model extracts a set of important features for small-business credit scoring for the observed sample, emphasizing credit programme characteristics, as well as entrepreneur's personal and business characteristics as the most important ones. Copyright 2005 John Wiley & Sons, Ltd. [source]


    Stochastic stability of a neural-net robot controller subject to signal-dependent noise in the learning rule

    INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING, Issue 6 2010
    Abraham K. Ishihara
    Abstract We consider a neural network-based controller for a rigid serial link manipulator with uncertain plant parameters. We assume that the training signal to the network is corrupted by signal-dependent noise. A radial basis function network is utilized in the feedforward control to approximate the unknown inverse dynamics. The weights are adaptively adjusted according to a gradient descent plus a regulation term (Narendra's e -modification). We prove a theorem that extends the Yoshizawa D-boundedness results to the stochastic setting. As in the deterministic setting, this result is particularly useful for neural network robot control when there exists bounded torque disturbances and neural net approximation errors over a known compact set. Using this result, we establish bounds on the feedback gains and learning rate parameters that guarantee the origin of the closed-loop system is semi-globally, uniformly bounded in expected value. Copyright 2009 John Wiley & Sons, Ltd. [source]


    Neural network-based adaptive attitude tracking control for flexible spacecraft with unknown high-frequency gain

    INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL PROCESSING, Issue 6 2010
    Qinglei Hu
    Abstract Adaptive control design using neural networks (a) is investigated for attitude tracking and vibration stabilization of a flexible spacecraft, which is operated at highly nonlinear dynamic regimes. The spacecraft considered consists of a rigid body and two flexible appendages, and it is assumed that the system parameters are unknown and the truncated model of the spacecraft has finite but arbitrary dimension as well, for the purpose of design. Based on this nonlinear model, the derivation of an adaptive control law using neural networks (NNs) is treated, when the dynamics of unstructured and state-dependent nonlinear function are completely unknown. A radial basis function network that is used here for synthesizing the controller and adaptive mechanisms is derived for adjusting the parameters of the network and estimating the unknown parameters. In this derivation, the Nussbaum gain technique is also employed to relax the sign assumption for the high-frequency gain for the neural adaptive control. Moreover, systematic design procedure is developed for the synthesis of adaptive NN tracking control with L2 -gain performance. The resulting closed-loop system is proven to be globally stable by Lyapunov's theory and the effect of the external disturbances and elastic vibrations on the tracking error can be attenuated to the prescribed level by appropriately choosing the design parameters. Numerical simulations are performed to show that attitude tracking control and vibration suppression are accomplished in spite of the presence of disturbance torque/parameter uncertainty. Copyright 2009 John Wiley & Sons, Ltd. [source]


    Fault detection and isolation in robotic manipulators via neural networks: A comparison among three architectures for residual analysis

    JOURNAL OF FIELD ROBOTICS (FORMERLY JOURNAL OF ROBOTIC SYSTEMS), Issue 7 2001
    Marco Henrique Terra
    In this article we discuss artificial neural networks-based fault detection and isolation (FDI) applications for robotic manipulators. The artificial neural networks (ANNs) are used for both residual generation and residual analysis. A multilayer perceptron (MLP) is employed to reproduce the dynamics of the robotic manipulator. Its outputs are compared with actual position and velocity measurements, generating the so-called residual vector. The residuals, when properly analyzed, provides an indication of the status of the robot (normal or faulty operation). Three ANNs architectures are employed in the residual analysis. The first is a radial basis function network (RBFN) which uses the residuals of position and velocity to perform fault identification. The second is again an RBFN, except that it uses only the velocity residuals. The third is an MLP which also performs fault identification utilizing only the velocity residuals. The MLP is trained with the classical back-propagation algorithm and the RBFN is trained with a Kohonen self-organizing map (KSOM). We validate the concepts discussed in a thorough simulation study of a Puma 560 and with experimental results with a 3-joint planar manipulator. 2001 John Wiley & Sons, Inc. [source]


    Rapid Determination of Invert Cane Sugar Adulteration in Honey Using FTIR Spectroscopy and Multivariate Analysis

    JOURNAL OF FOOD SCIENCE, Issue 6 2003
    J. Irudayaraj
    ABSTRACT: Fourier transform infrared spectroscopy with an attenuated total reflection sampling accessory was combined with multivariate analysis to determine the level (1% to 25%, wt/wt) of invert cane sugar adulteration in honey. On the basis of the spectral data compression by principal component analysis and partial least squares, linear discriminant analysis (LDA), and canonical variate analysis (CVA), models were developed and validated. Two types of artificial neural networks were applied: a quick back propagation network (BPN) and a radial basis function network (RBFN). The prediction success rates were better with LDA (93.75% for validation set) and BPN (93.75%) than with CVA (87.50%) and RBFN (81.25%). [source]


    A Cartesian-grid collocation technique with integrated radial basis functions for mixed boundary value problems

    INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN ENGINEERING, Issue 4 2010
    Phong B. H. Le
    Abstract In this paper, high-order systems are reformulated as first-order systems, which are then numerically solved by a collocation method. The collocation method is based on Cartesian discretization with 1D-integrated radial basis function networks (1D-IRBFN) (Numer. Meth. Partial Differential Equations 2007; 23:1192,1210). The present method is enhanced by a new boundary interpolation technique based on 1D-IRBFN, which is introduced to obtain variable approximation at irregular points in irregular domains. The proposed method is well suited to problems with mixed boundary conditions on both regular and irregular domains. The main results obtained are (a) the boundary conditions for the reformulated problem are of Dirichlet type only; (b) the integrated RBFN approximation avoids the well-known reduction of convergence rate associated with differential formulations; (c) the primary variable (e.g. displacement, temperature) and the dual variable (e.g. stress, temperature gradient) have similar convergence order; (d) the volumetric locking effects associated with incompressible materials in solid mechanics are alleviated. Numerical experiments show that the proposed method achieves very good accuracy and high convergence rates. Copyright 2009 John Wiley & Sons, Ltd. [source]


    A Cartesian grid technique based on one-dimensional integrated radial basis function networks for natural convection in concentric annuli

    INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS, Issue 12 2008
    N. Mai-Duy
    Abstract This paper reports a radial basis function (RBF)-based Cartesian grid technique for the simulation of two-dimensional buoyancy-driven flow in concentric annuli. The continuity and momentum equations are represented in the equivalent stream function formulation that reduces the number of equations from three to one, but involves higher-order derivatives. The present technique uses a Cartesian grid to discretize the problem domain. Along a grid line, one-dimensional integrated RBF networks (1D-IRBFNs) are employed to represent the field variables. The capability of 1D-IRBFNs to handle unstructured points with accuracy is exploited to describe non-rectangular boundaries in a Cartesian grid, while the method's ability to avoid the reduction of convergence rate caused by differentiation is instrumental in improving the quality of the approximation of higher-order derivatives. The method is applied to simulate thermally driven flows in annuli between two circular cylinders and between an outer square cylinder and an inner circular cylinder. High Rayleigh number solutions are achieved and they are in good agreement with previously published numerical data. Copyright 2007 John Wiley & Sons, Ltd. [source]


    Computing non-Newtonian fluid flow with radial basis function networks

    INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS, Issue 12 2005
    N. Mai-Duy
    Abstract This paper is concerned with the application of radial basis function networks (RBFNs) for solving non-Newtonian fluid flow problems. Indirect RBFNs, which are based on an integration process, are employed to represent the solution variables; the governing differential equations are discretized by means of point collocation. To enhance numerical stability, stress-splitting techniques are utilized. The proposed method is verified through the computation of the rectilinear and non-rectilinear flows in a straight duct and the axisymmetric flow in an undulating tube using Newtonian, power-law, Criminale,Ericksen,Filbey (CEF) and Oldroyd-B models. The obtained results are in good agreement with the analytic and benchmark solutions. Copyright 2005 John Wiley & Sons, Ltd. [source]


    Inexact information aided, low-cost, distributed genetic algorithms for aerodynamic shape optimization

    INTERNATIONAL JOURNAL FOR NUMERICAL METHODS IN FLUIDS, Issue 10-11 2003
    Marios K. Karakasis
    Abstract Despite its robustness, the design and optimization of aerodynamic shapes using genetic algorithms suffers from high computing cost requirements, due to excessive calls to Computational Fluid Dynamics tools for the evaluation of candidate solutions. To alleviate this problem, either the use of distributed genetic algorithms or the implementation of surrogate evaluation models have separately been proposed in the past. A distributed genetic algorithm relies on the handling of population subsets that evolve in a semi-isolated manner by regularly exchanging their best individuals. It is known that distributed schemes generally outperform single-population ones. On the other hand, the implementation of less costly surrogate evaluation tools, such as the autocatalytic radial basis function networks developed by the authors for the purpose of getting rid of most of the ,useless' exact evaluations, reduces considerably the computational cost. The aim of the present paper is to employ a surrogate evaluation model in the context of a distributed genetic algorithm and to demonstrate that the combination of both results in maximum economy in CPU cost. In addition, whenever a multiprocessor system is available, the gain is much more pronounced, since the new optimization method maximizes parallel efficiency. The proposed method is used to solve inverse design and optimization problems in aeronautics and turbomachinery. Copyright 2003 John Wiley & Sons, Ltd. [source]


    Neural network model for the efficient calculation of Green's functions in layered media

    INTERNATIONAL JOURNAL OF RF AND MICROWAVE COMPUTER-AIDED ENGINEERING, Issue 2 2003
    E. A. Soliman
    Abstract In this article, neural networks are employed for fast and efficient calculation of Green's functions in a layered medium. Radial basis function networks (RBFNs) are effectively trained to estimate the coefficients and the exponents that represent a Green's function in the discrete complex image method (DCIM). Results show very good agreement with the DCIM, and the trained RBFNs are very fast compared with the corresponding DCIM. 2003 Wiley Periodicals, Inc. Int J RF and Microwave CAE 13: 128,135, 2003. [source]