TY - JOUR

T1 - Localized bases of eigensubspaces and operator compression

AU - Weinan, E.

AU - Li, Tiejun

AU - Lu, Jianfeng

N1 - Copyright:
Copyright 2010 Elsevier B.V., All rights reserved.

PY - 2010/1/26

Y1 - 2010/1/26

N2 - Given a complex local operator, such as the generator of a Markov chain on a large network, a differential operator, or a large sparse matrix that comes fromthe discretization of a differential operator, we would like to find its best finite dimensional approximation with a given dimension. The answer to this question is often given simply by the projection of the original operator to its eigensubspace of the given dimension that corresponds to the smallest or largest eigenvalues, depending on the setting. The representation of such subspaces, however, is far from being unique and our interest is to find the most localized bases for these subspaces. The reduced operator using these bases would have sparsity features similar to that of the original operator. We will discuss different ways of obtaining localized bases, and we will give an explicit characterization of the decay rate of these basis functions. We will also discuss efficient numerical algorithms for finding such basis functions and the reduced (or compressed) operator.

AB - Given a complex local operator, such as the generator of a Markov chain on a large network, a differential operator, or a large sparse matrix that comes fromthe discretization of a differential operator, we would like to find its best finite dimensional approximation with a given dimension. The answer to this question is often given simply by the projection of the original operator to its eigensubspace of the given dimension that corresponds to the smallest or largest eigenvalues, depending on the setting. The representation of such subspaces, however, is far from being unique and our interest is to find the most localized bases for these subspaces. The reduced operator using these bases would have sparsity features similar to that of the original operator. We will discuss different ways of obtaining localized bases, and we will give an explicit characterization of the decay rate of these basis functions. We will also discuss efficient numerical algorithms for finding such basis functions and the reduced (or compressed) operator.

KW - Singular value decomposition

KW - Subspace iteration

UR - http://www.scopus.com/inward/record.url?scp=76549093394&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=76549093394&partnerID=8YFLogxK

U2 - 10.1073/pnas.0913345107

DO - 10.1073/pnas.0913345107

M3 - Article

C2 - 20080703

AN - SCOPUS:76549093394

VL - 107

SP - 1273

EP - 1278

JO - Proceedings of the National Academy of Sciences of the United States of America

JF - Proceedings of the National Academy of Sciences of the United States of America

SN - 0027-8424

IS - 4

ER -