Radial basis functions in the Fourier domain
نویسنده
چکیده
We demonstrate a relationship between the singular values of the design matrix and the discrete Fourier transform of the radial function for radial basis function networks. We then show how regularisation leads to high frequency filtering of the network output. In certain circumstances, this allows the network parameters to be chosen a priori to appropriately bias the learning process.
منابع مشابه
Time-Discontinuous Finite Element Analysis of Two-Dimensional Elastodynamic Problems using Complex Fourier Shape Functions
This paper reformulates a time-discontinuous finite element method (TD-FEM) based on a new class of shape functions, called complex Fourier hereafter, for solving two-dimensional elastodynamic problems. These shape functions, which are derived from their corresponding radial basis functions, have some advantages such as the satisfaction of exponential and trigonometric function fields in comple...
متن کاملNumerical Simulation of 1D Linear Telegraph Equation With Variable Coefficients Using Meshless Local Radial Point Interpolation (MLRPI)
In the current work, we implement the meshless local radial point interpolation (MLRPI) method to find numerical solution of one-dimensional linear telegraph equations with variable coefficients. The MLRPI method, as a meshless technique, does not require any background integration cells and all integrations are carried out locally over small quadrature domains of regular shapes, such as lines ...
متن کاملA numerical solution of mixed Volterra Fredholm integral equations of Urysohn type on non-rectangular regions using meshless methods
In this paper, we propose a new numerical method for solution of Urysohn two dimensional mixed Volterra-Fredholm integral equations of the second kind on a non-rectangular domain. The method approximates the solution by the discrete collocation method based on inverse multiquadric radial basis functions (RBFs) constructed on a set of disordered data. The method is a meshless method, because it ...
متن کاملA meshless method for optimal control problem of Volterra-Fredholm integral equations using multiquadratic radial basis functions
In this paper, a numerical method is proposed for solving optimal control problem of Volterra integral equations using radial basis functions (RBFs) for approximating unknown function. Actually, the method is based on interpolation by radial basis functions including multiquadrics (MQs), to determine the control vector and the corresponding state vector in linear dynamic system while minimizing...
متن کاملThe method of radial basis functions for the solution of nonlinear Fredholm integral equations system.
In this paper, An effective and simple numerical method is proposed for solving systems of integral equations using radial basis functions (RBFs). We present an algorithm based on interpolation by radial basis functions including multiquadratics (MQs), using Legendre-Gauss-Lobatto nodes and weights. Also a theorem is proved for convergence of the algorithm. Some numerical examples are presented...
متن کامل