An Adaptive Pruning Algorithm for the Discrete L-Curve Criterion

Publication: ResearchReport – Annual report year: 2004

Standard

An Adaptive Pruning Algorithm for the Discrete L-Curve Criterion. / Hansen, Per Christian; Jensen, Toke Koldborg; Rodriguez, Giuseppe.

2004. 23 p.

Publication: ResearchReport – Annual report year: 2004

Harvard

APA

CBE

MLA

Vancouver

Author

Hansen, Per Christian; Jensen, Toke Koldborg; Rodriguez, Giuseppe / An Adaptive Pruning Algorithm for the Discrete L-Curve Criterion.

2004. 23 p.

Publication: ResearchReport – Annual report year: 2004

Bibtex

@book{67dac088b5cd4bfeb523df50f698e54e,
title = "An Adaptive Pruning Algorithm for the Discrete L-Curve Criterion",
keywords = "L-curve criterion, parameter-choice method., regularization, Discrete ill-posed problems",
author = "Hansen, {Per Christian} and Jensen, {Toke Koldborg} and Giuseppe Rodriguez",
year = "2004",

}

RIS

TY - RPRT

T1 - An Adaptive Pruning Algorithm for the Discrete L-Curve Criterion

A1 - Hansen,Per Christian

A1 - Jensen,Toke Koldborg

A1 - Rodriguez,Giuseppe

AU - Hansen,Per Christian

AU - Jensen,Toke Koldborg

AU - Rodriguez,Giuseppe

PY - 2004

Y1 - 2004

N2 - We describe a robust and adaptive implementation of the L-curve criterion, i.e., for locating the corner of a discrete L-curve consisting of a log-log plot of corresponding residual and solution norms of regularized solutions from a method with a discrete regularization parameter (such as truncated SVD or regularizing CG iterations). Our algorithm needs no pre-defined parameters, and in order to capture the global features of the curve in an adaptive fashion, we use a sequence of pruned L-curves that correspond to considering the curves at different scales. We compare our new algorithm to existing algoritms and demonstrate its robustness by numerical examples.

AB - We describe a robust and adaptive implementation of the L-curve criterion, i.e., for locating the corner of a discrete L-curve consisting of a log-log plot of corresponding residual and solution norms of regularized solutions from a method with a discrete regularization parameter (such as truncated SVD or regularizing CG iterations). Our algorithm needs no pre-defined parameters, and in order to capture the global features of the curve in an adaptive fashion, we use a sequence of pruned L-curves that correspond to considering the curves at different scales. We compare our new algorithm to existing algoritms and demonstrate its robustness by numerical examples.

KW - L-curve criterion

KW - parameter-choice method.

KW - regularization

KW - Discrete ill-posed problems

UR - http://www.imm.dtu.dk/pubdb/p.php?3381

BT - An Adaptive Pruning Algorithm for the Discrete L-Curve Criterion

ER -