Search results for "Basis function"
showing 10 items of 103 documents
The integral‐direct coupled cluster singles and doubles model
1996
An efficient and highly vectorized implementation of the coupled cluster singles and doubles (CCSD) model using a direct atomic integral technique is presented. The minimal number of n6processes has been implemented for the most time consuming terms and point group symmetry is used to further reduce operation counts and memory requirements. The significantly increased application range of the CCSD method is illustrated with sample calculations on several systems with more than 500 basis functions. Furthermore, we present the basic trends of an open ended algorithm and discuss the use of integral prescreening. © 1996 American Institute of Physics.
Fake Nodes approximation for Magnetic Particle Imaging
2020
Accurately reconstructing functions with discontinuities is the key tool in many bio-imaging applications as, for instance, in Magnetic Particle Imaging (MPI). In this paper, we apply a method for scattered data interpolation, named mapped bases or Fake Nodes approach, which incorporates discontinuities via a suitable mapping function. This technique naturally mitigates the Gibbs phenomenon, as numerical evidence for reconstructing MPI images confirms.
Experimental validation for spectrum cartography using adaptive multi-kernels
2017
This paper validates the functionality of an algorithm for spectrum cartography, generating a radio environment map (REM) using adaptive radial basis functions (RBF) based on a limited number of measurements. The power at all locations is estimated as a linear combination of different RBFs without assuming any prior information about either power spectral densities (PSD) of the transmitters or their locations. The RBFs are represented as centroids at optimized locations, using machine learning to jointly optimize their positions, weights and Gaussian decaying parameters. Optimization is performed using expectation maximization with a least squares loss function and a quadratic regularizer. …
Structured Output SVM for Remote Sensing Image Classification
2011
Traditional kernel classifiers assume independence among the classification outputs. As a consequence, each misclassification receives the same weight in the loss function. Moreover, the kernel function only takes into account the similarity between input values and ignores possible relationships between the classes to be predicted. These assumptions are not consistent for most of real-life problems. In the particular case of remote sensing data, this is not a good assumption either. Segmentation of images acquired by airborne or satellite sensors is a very active field of research in which one tries to classify a pixel into a predefined set of classes of interest (e.g. water, grass, trees,…
Support Vector Machines for Crop Classification Using Hyperspectral Data
2003
In this communication, we propose the use of Support Vector Machines (SVM) for crop classification using hyperspectral images. SVM are benchmarked to well–known neural networks such as multilayer perceptrons (MLP), Radial Basis Functions (RBF) and Co-Active Neural Fuzzy Inference Systems (CANFIS). Models are analyzed in terms of efficiency and robustness, which is tested according to their suitability to real–time working conditions whenever a preprocessing stage is not possible. This can be simulated by considering models with and without a preprocessing stage. Four scenarios (128, 6, 3 and 2 bands) are thus evaluated. Several conclusions are drawn: (1) SVM yield better outcomes than neura…
Sequential Learning with LS-SVM for Large-Scale Data Sets
2006
We present a subspace-based variant of LS-SVMs (i.e. regularization networks) that sequentially processes the data and is hence especially suited for online learning tasks. The algorithm works by selecting from the data set a small subset of basis functions that is subsequently used to approximate the full kernel on arbitrary points. This subset is identified online from the data stream. We improve upon existing approaches (esp. the kernel recursive least squares algorithm) by proposing a new, supervised criterion for the selection of the relevant basis functions that takes into account the approximation error incurred from approximating the kernel as well as the reduction of the cost in th…
Implementation of pattern recognition algorithm based on RBF neural network
2002
In this paper, we present implementations of a pattern recognition algorithm which uses a RBF (Radial Basis Function) neural network. Our aim is to elaborate a quite efficient system which realizes real time faces tracking and identity verification in natural video sequences. Hardware implementations have been realized on an embedded system developed by our laboratory. This system is based on a DSP (Digital Signal Processor) TMS320C6x. The optimization of implementations allow us to obtain a processing speed of 4.8 images (240x320 pixels) per second with a correct rate of 95% of faces tracking and identity verification.
An Advanced Numerical Model in Solving Thin-Wire Integral Equations by Using Semi-Orthogonal Compactly Supported Spline Wavelets
2003
Abstract—In this paper, the semi-orthogonal compactly sup- ported spline wavelets are used as basis functions for the efficient solution of the thin-wire electric field integral equation (EFIE) in frequency domain. The method of moments (MoM) is used via the Galerkin procedure. Conventional MoM directly applied to the EFIE, leads to dense matrix which often becomes computation- ally intractable when large-scale problems are approached. To overcome these difficulties, wavelets can be used as a basis set so obtaining the generation of a sparse matrix; this is due to the local supports and the vanishing moments properties of the wavelets. In the paper, this technique is applied to analyze elec…
Highly efficient full-wave electromagnetic analysis of 3-D arbitrarily shaped waveguide microwave devices using an integral equation technique
2015
A novel technique for the full-wave analysis of 3-D complex waveguide devices is presented. This new formulation, based on the Boundary Integral-Resonant Mode Expansion (BI-RME) method, allows the rigorous full-wave electromagnetic characterization of 3-D arbitrarily shaped metallic structures making use of extremely low CPU resources (both time and memory). The unknown electric current density on the surface of the metallic elements is represented by means of Rao-Wilton-Glisson basis functions, and an algebraic procedure based on a singular value decomposition is applied to transform such functions into the classical solenoidal and nonsolenoidal basis functions needed by the original BI-RM…
Structural Health Monitoring Procedure for Composite Structures through the use of Artificial Neural Networks
2015
In this paper different architectures of Artificial Neural Networks (ANNs) for structural damage detection are studied. The main objective is to investigate an ANN able to detect and localize damage without any prior knowledge on its characteristics so as to serve as a real-time data processor for Structural Health Monitoring (SHM) systems. Two different architectures are studied: the standard feed-forward Multi Layer Perceptron (MLP) and the Radial Basis Function (RBF) ANNs. The training data are given, in terms of a Damage Index ℑD, properly defined using a piezoelectric sensor signal output to obtain suitable information on the damage position and dimensions. The electromechanical respon…