An Improved Vectorial Kernel Orthogonal Greedy Algorithm
D. Wirtz, and B. Haasdonk. SimTech Preprint, University of Stuttgart, (2012)
Abstract
This work is concerned with derivation and analysis of a modifed vectorial
kernel orthogonal greedy algorithm (VKOGA) for approximation of nonlinear
vectorial functions. The algorithm pursues simultaneous approximation
of all vector components over a shared linear subspace of the underlying
function Hilbert space in a greedy fashion 14, 33 and inherits
the selection principle of the f=P-Greedy algorithm 18. For the
considered algorithm we perform a limit analysis of the selection
criteria for already included subspace basis functions. We show that
the approximation gain is bounded globally and for the multivariate
case the limit functions correspond to a directional Hermite interpolation.
We further prove algebraic convergence similar to 13, improved
by a dimension-dependent factor, and introduce a new a-posteriori
error bound. Comparison to related variants of our algorithm are
presented. Targeted applications of this algorithm are model reduction
of multiscale models 40.
%0 Report
%1 Wirtz2012
%A Wirtz, Daniel
%A Haasdonk, Bernard
%D 2012
%K anm ians imported
%T An Improved Vectorial Kernel Orthogonal Greedy Algorithm
%U http://www.simtech.uni-stuttgart.de/publikationen/prints.php?ID=742
%X This work is concerned with derivation and analysis of a modifed vectorial
kernel orthogonal greedy algorithm (VKOGA) for approximation of nonlinear
vectorial functions. The algorithm pursues simultaneous approximation
of all vector components over a shared linear subspace of the underlying
function Hilbert space in a greedy fashion 14, 33 and inherits
the selection principle of the f=P-Greedy algorithm 18. For the
considered algorithm we perform a limit analysis of the selection
criteria for already included subspace basis functions. We show that
the approximation gain is bounded globally and for the multivariate
case the limit functions correspond to a directional Hermite interpolation.
We further prove algebraic convergence similar to 13, improved
by a dimension-dependent factor, and introduce a new a-posteriori
error bound. Comparison to related variants of our algorithm are
presented. Targeted applications of this algorithm are model reduction
of multiscale models 40.
@techreport{Wirtz2012,
abstract = {This work is concerned with derivation and analysis of a modifed vectorial
kernel orthogonal greedy algorithm (VKOGA) for approximation of nonlinear
vectorial functions. The algorithm pursues simultaneous approximation
of all vector components over a shared linear subspace of the underlying
function Hilbert space in a greedy fashion [14, 33] and inherits
the selection principle of the f=P-Greedy algorithm [18]. For the
considered algorithm we perform a limit analysis of the selection
criteria for already included subspace basis functions. We show that
the approximation gain is bounded globally and for the multivariate
case the limit functions correspond to a directional Hermite interpolation.
We further prove algebraic convergence similar to [13], improved
by a dimension-dependent factor, and introduce a new a-posteriori
error bound. Comparison to related variants of our algorithm are
presented. Targeted applications of this algorithm are model reduction
of multiscale models [40].},
added-at = {2021-09-29T14:33:27.000+0200},
author = {Wirtz, Daniel and Haasdonk, Bernard},
biburl = {https://puma.ub.uni-stuttgart.de/bibtex/20b683f799d0ffc1ab4766803d265c8e6/britsteiner},
file = {:PDF/WH12c_preprint.pdf:PDF},
groups = {Greedy},
institution = {University of Stuttgart},
interhash = {40a8fffd6725cda447e58936382e62b2},
intrahash = {0b683f799d0ffc1ab4766803d265c8e6},
keywords = {anm ians imported},
note = {{I}n preparation},
owner = {haasdonk},
timestamp = {2021-09-29T12:35:04.000+0200},
title = {An Improved Vectorial Kernel Orthogonal Greedy Algorithm},
type = {SimTech Preprint},
url = {http://www.simtech.uni-stuttgart.de/publikationen/prints.php?ID=742},
year = 2012
}