Search results for "Sparse matrix"
showing 10 items of 23 documents
Wavelet-like bases for thin-wire integral equations in electromagnetics
2005
AbstractIn this paper, wavelets are used in solving, by the method of moments, a modified version of the thin-wire electric field integral equation, in frequency domain. The time domain electromagnetic quantities, are obtained by using the inverse discrete fast Fourier transform. The retarded scalar electric and vector magnetic potentials are employed in order to obtain the integral formulation. The discretized model generated by applying the direct method of moments via point-matching procedure, results in a linear system with a dense matrix which have to be solved for each frequency of the Fourier spectrum of the time domain impressed source. Therefore, orthogonal wavelet-like basis trans…
Scatter search for the profile minimization problem
2014
We study the problem of minimizing the profile of a graph and develop a solution method by following the tenets of scatter search. Our procedure exploits the network structure of the problem and includes strategies that produce a computationally efficient and agile search. Among several mechanisms, our search includes path relinking as the basis for combining solutions to generate new ones. The profile minimization problem PMP is NP-Hard and has relevant applications in numerical analysis techniques that rely on manipulating large sparse matrices. The problem was proposed in the early 1970s but the state-of-the-art does not include a method that could be considered powerful by today's compu…
Reducing the bandwidth of a sparse matrix with tabu search
2001
The bandwidth of a matrix { } ij a A = is defined as the maximum absolute difference between i and j for which 0 ≠ ij a . The problem of reducing the bandwidth of a matrix consists of finding a permutation of the rows and columns that keeps the nonzero elements in a band that is as close as possible to the main diagonal of the matrix. This NP-complete problem can also be formulated as a labeling of vertices on a graph, where edges are the nonzero elements of the corresponding symmetrical matrix. Many bandwidth reduction algorithms have been developed since the 1960s and applied to structural engineering, fluid dynamics and network analysis. For the most part, these procedures do not incorpo…
A Linear Cost Algorithm to Compute the Discrete Gabor Transform
2010
In this paper, we propose an alternative efficient method to calculate the Gabor coefficients of a signal given a synthesis window with a support of size much lesser than the length of the signal. The algorithm uses the canonical dual of the window (which does not need to be calculated beforehand) and achieves a computational cost that is linear with the signal length in both analysis and synthesis. This is done by exploiting the block structure of the matrices and using an ad hoc Cholesky decomposition of the Gabor frame matrix.
Multi-label classification using boolean matrix decomposition
2012
This paper introduces a new multi-label classifier based on Boolean matrix decomposition. Boolean matrix decomposition is used to extract, from the full label matrix, latent labels representing useful Boolean combinations of the original labels. Base level models predict latent labels, which are subsequently transformed into the actual labels by Boolean matrix multiplication with the second matrix from the decomposition. The new method is tested on six publicly available datasets with varying numbers of labels. The experimental evaluation shows that the new method works particularly well on datasets with a large number of labels and strong dependencies among them.
Information Dynamics Analysis: A new approach based on Sparse Identification of Linear Parametric Models*
2020
The framework of information dynamics allows to quantify different aspects of the statistical structure of multivariate processes reflecting the temporal dynamics of a complex network. The information transfer from one process to another can be quantified through Transfer Entropy, and under the assumption of joint Gaussian variables it is strictly related to the concept of Granger Causality (GC). According to the most recent developments in the field, the computation of GC entails representing the processes through a Vector Autoregressive (VAR) model and a state space (SS) model typically identified by means of the Ordinary Least Squares (OLS). In this work, we propose a new identification …
Propagation pattern analysis during atrial fibrillation based on the adaptive group LASSO.
2012
The present study introduces sparse modeling for the estimation of propagation patterns in intracardiac atrial fibrillation (AF) signals. The estimation is based on the partial directed coherence (PDC) function, derived from fitting a multivariate autoregressive model to the observed signals. A sparse optimization method is proposed for estimation of the model parameters, namely, the adaptive group least absolute selection and shrinkage operator (aLASSO). In simulations aLASSO was found superior to the commonly used least-squares (LS) estimation with respect to estimation performance. The normalized error between the true and estimated model parameters dropped from 0.200.04 for LS estimatio…
Numerical Study of Two Sparse AMG-methods
2003
A sparse algebraic multigrid method is studied as a cheap and accurate way to compute approximations of Schur complements of matrices arising from the discretization of some symmetric and positive definite partial differential operators. The construction of such a multigrid is discussed and numerical experiments are used to verify the properties of the method.
Phase retrieval of a Kolmogorov phase screen from very sparse data using four binary masks
2020
We investigate experimentally the phase retrieval of a Kolmogorov phase screen from very sparse data by modulating its amplitude with four binary masks and compare the retrieved phase screen to the ground truth measured with a surface profiler. Previously, we have shown in simulations that this kind of modulation can be successfully used for the phase retrieval of a Kolmogorov phase screen. After subtracting the ground truth from the retrieved phase screen, the root-mean-square error decreased from 0.14 µm to 0.10 µm. We conclude that a Kolmogorov phase screen can be recovered using simple modulation and very sparse data.
Method specific Cholesky decomposition : Coulomb and exchange energies
2008
We present a novel approach to the calculation of the Coulomb and exchange contributions to the total electronic energy in self consistent field and density functional theory. The numerical procedure is based on the Cholesky decomposition and involves decomposition of specific Hadamard product matrices that enter the energy expression. In this way, we determine an auxiliary basis and obtain a dramatic reduction in size as compared to the resolution of identity (RI) method. Although the auxiliary basis is determined from the energy expression, we have complete control of the errors in the gradient or Fock matrix. Another important advantage of this method specific Cholesky decomposition is t…