Comparing RSVD and Krylov methods for linear inverse problems

In this work we address regularization parameter estimation for ill-posed linear inverse problems with an ℓ2 penalty. Regularization parameter selection is of utmost importance for all of inverse problems and estimating it generally relies on the experience of the practitioner. For regularization wi...

Full description

Saved in:
Bibliographic Details
Published in:Computers & geosciences Vol. 137; p. 104427
Main Authors: Luiken, Nick, van Leeuwen, Tristan
Format: Journal Article
Language:English
Published: Elsevier Ltd 01.04.2020
Subjects:
ISSN:0098-3004, 1873-7803
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract In this work we address regularization parameter estimation for ill-posed linear inverse problems with an ℓ2 penalty. Regularization parameter selection is of utmost importance for all of inverse problems and estimating it generally relies on the experience of the practitioner. For regularization with an ℓ2 penalty there exist a lot of parameter selection methods that exploit the fact that the solution and the residual can be written in explicit form. Parameter selection methods are functionals that depend on the regularization parameter where the minimizer is the desired regularization parameter that should lead to a good solution. Evaluation of these parameter selection methods still requires solving the inverse problem multiple times. Efficient evaluation of the parameter selection methods can be done through model order reduction. Two popular model order reduction techniques are Lanczos based methods (a Krylov subspace method) and the Randomized Singular Value Decomposition (RSVD). In this work we compare the two approaches. We derive error bounds for the parameter selection methods using the RSVD. We compare the performance of the Lanczos process versus the performance of RSVD for efficient parameter selection. The RSVD algorithm we use is based on the Adaptive Randomized Range Finder algorithm which allows for easy determination of the dimension of the reduced order model. Some parameter selection also require the evaluation of the trace of a large matrix. We compare the use of a randomized trace estimator versus the use of the Ritz values from the Lanczos process. The examples we use for our experiments are two model problems from geosciences. •Overview of some parameter selection methods with references to the relevant literature.•Comparison of Krylov based methods versus the RSVD for parameter selection.•Error bounds for TSVD and RSVD for parameter selection methods.•Comparison of trace estimators for different types of inverse problems.
AbstractList In this work we address regularization parameter estimation for ill-posed linear inverse problems with an ℓ2 penalty. Regularization parameter selection is of utmost importance for all of inverse problems and estimating it generally relies on the experience of the practitioner. For regularization with an ℓ2 penalty there exist a lot of parameter selection methods that exploit the fact that the solution and the residual can be written in explicit form. Parameter selection methods are functionals that depend on the regularization parameter where the minimizer is the desired regularization parameter that should lead to a good solution. Evaluation of these parameter selection methods still requires solving the inverse problem multiple times. Efficient evaluation of the parameter selection methods can be done through model order reduction. Two popular model order reduction techniques are Lanczos based methods (a Krylov subspace method) and the Randomized Singular Value Decomposition (RSVD). In this work we compare the two approaches. We derive error bounds for the parameter selection methods using the RSVD. We compare the performance of the Lanczos process versus the performance of RSVD for efficient parameter selection. The RSVD algorithm we use is based on the Adaptive Randomized Range Finder algorithm which allows for easy determination of the dimension of the reduced order model. Some parameter selection also require the evaluation of the trace of a large matrix. We compare the use of a randomized trace estimator versus the use of the Ritz values from the Lanczos process. The examples we use for our experiments are two model problems from geosciences. •Overview of some parameter selection methods with references to the relevant literature.•Comparison of Krylov based methods versus the RSVD for parameter selection.•Error bounds for TSVD and RSVD for parameter selection methods.•Comparison of trace estimators for different types of inverse problems.
In this work we address regularization parameter estimation for ill-posed linear inverse problems with an ℓ2 penalty. Regularization parameter selection is of utmost importance for all of inverse problems and estimating it generally relies on the experience of the practitioner. For regularization with an ℓ2 penalty there exist a lot of parameter selection methods that exploit the fact that the solution and the residual can be written in explicit form. Parameter selection methods are functionals that depend on the regularization parameter where the minimizer is the desired regularization parameter that should lead to a good solution. Evaluation of these parameter selection methods still requires solving the inverse problem multiple times. Efficient evaluation of the parameter selection methods can be done through model order reduction. Two popular model order reduction techniques are Lanczos based methods (a Krylov subspace method) and the Randomized Singular Value Decomposition (RSVD). In this work we compare the two approaches. We derive error bounds for the parameter selection methods using the RSVD. We compare the performance of the Lanczos process versus the performance of RSVD for efficient parameter selection. The RSVD algorithm we use is based on the Adaptive Randomized Range Finder algorithm which allows for easy determination of the dimension of the reduced order model. Some parameter selection also require the evaluation of the trace of a large matrix. We compare the use of a randomized trace estimator versus the use of the Ritz values from the Lanczos process. The examples we use for our experiments are two model problems from geosciences.
ArticleNumber 104427
Author van Leeuwen, Tristan
Luiken, Nick
Author_xml – sequence: 1
  givenname: Nick
  surname: Luiken
  fullname: Luiken, Nick
  email: n.a.luiken@uu.nl
– sequence: 2
  givenname: Tristan
  surname: van Leeuwen
  fullname: van Leeuwen, Tristan
  email: t.vanleeuwen@uu.nl
BookMark eNqFkD1PwzAQhi1UJNrCL2DxyJJiO07iDB1Q-RSVkPhaLcc5F1eJXey0Uv89KWFigOmk0_u8p3smaOS8A4TOKZlRQvPL9UyrFfgZI-yw4ZwVR2hMRZEmhSDpCI0JKUWSEsJP0CTGNSGEMZGN0Xzh240K1q3w88v7NVauxo9h3_gdbqH78HXExgfcWAcqYOt2ECLgTfBVA208RcdGNRHOfuYUvd3evC7uk-XT3cPiapkoTniXZLwuCs2B8oqVRhBt8rqgOk9Bl8Ao45VRvKxKmmWQGUYFp1DnFVdKZAbKIp2ii6G3P_y5hdjJ1kYNTaMc-G2UjPefZQUXtI-mQ1QHH2MAIzfBtirsJSXyIEuu5bcseZAlB1k9Vf6itO1UZ73rgrLNP-x8YKE3sLMQZNQWnIbaBtCdrL39k_8Cmy-HdQ
CitedBy_id crossref_primary_10_1088_1361_6420_ac245d
crossref_primary_10_1109_ACCESS_2022_3174208
Cites_doi 10.1137/090771806
10.1016/0041-5553(84)90253-2
10.1016/j.cageo.2007.02.003
10.1080/00401706.1979.10489751
10.1137/0914086
10.1190/geo2017-0386.1
10.1023/A:1022383005969
10.1137/0904012
10.1137/S0895479899345960
10.1088/0266-5611/24/3/034006
10.1007/BF01937276
10.1016/j.cageo.2018.09.005
10.1016/j.matcom.2011.01.016
10.1088/0266-5611/22/5/021
10.1088/0266-5611/29/8/085008
10.1080/10618600.1997.10474725
10.1016/0377-0427(96)00018-0
10.1080/03610919008812866
10.1137/16M1104974
10.1137/15M1030200
10.1137/1034115
10.1016/S0098-3004(01)00009-7
10.1016/j.cam.2011.09.039
10.1137/S1064827593252672
10.1002/nla.802
ContentType Journal Article
Copyright 2020 Elsevier Ltd
Copyright_xml – notice: 2020 Elsevier Ltd
DBID AAYXX
CITATION
7S9
L.6
DOI 10.1016/j.cageo.2020.104427
DatabaseName CrossRef
AGRICOLA
AGRICOLA - Academic
DatabaseTitle CrossRef
AGRICOLA
AGRICOLA - Academic
DatabaseTitleList
AGRICOLA
DeliveryMethod fulltext_linktorsrc
Discipline Geology
EISSN 1873-7803
ExternalDocumentID 10_1016_j_cageo_2020_104427
S0098300418306952
GroupedDBID --K
--M
.DC
.~1
0R~
1B1
1RT
1~.
1~5
29F
4.4
457
4G.
5GY
5VS
7-5
71M
8P~
9JN
AABNK
AACTN
AAEDT
AAEDW
AAIAV
AAIKJ
AAKOC
AALRI
AAOAW
AAQFI
AAQXK
AAXUO
AAYFN
ABBOA
ABFNM
ABMAC
ABQEM
ABQYD
ABXDB
ABYKQ
ACDAQ
ACGFS
ACLVX
ACNNM
ACRLP
ACSBN
ACZNC
ADBBV
ADEZE
ADJOM
ADMUD
AEBSH
AEKER
AENEX
AFKWA
AFTJW
AGHFR
AGUBO
AGYEJ
AHHHB
AHZHX
AIALX
AIEXJ
AIKHN
AITUG
AJBFU
AJOXV
ALMA_UNASSIGNED_HOLDINGS
AMFUW
AMRAJ
AOUOD
ASPBG
ATOGT
AVWKF
AXJTR
AZFZN
BKOJK
BLXMC
CS3
DU5
EBS
EFJIC
EFLBG
EJD
EO8
EO9
EP2
EP3
F5P
FDB
FEDTE
FGOYB
FIRID
FNPLU
FYGXN
G-2
G-Q
GBLVA
GBOLZ
HLZ
HMA
HVGLF
HZ~
IHE
IMUCA
J1W
KOM
LG9
LY3
M41
MO0
N9A
O-L
O9-
OAUVE
OZT
P-8
P-9
P2P
PC.
Q38
R2-
RIG
ROL
RPZ
SBC
SDF
SDG
SDP
SEP
SES
SEW
SPC
SPCBC
SSE
SSV
SSZ
T5K
TN5
WUQ
ZCA
ZMT
~02
~G-
9DU
AAHBH
AATTM
AAXKI
AAYWO
AAYXX
ABJNI
ABWVN
ACLOT
ACRPL
ACVFH
ADCNI
ADNMO
ADXHL
AEIPS
AEUPX
AFJKZ
AFPUW
AGQPQ
AIGII
AIIUN
AKBMS
AKRWK
AKYEP
ANKPU
APXCP
CITATION
EFKBS
~HD
7S9
L.6
ID FETCH-LOGICAL-a404t-54d77c4e14b29f80cf6d71c63ec9e2124bfa49b9155e5f21841ed6b4aa85fe973
ISICitedReferencesCount 3
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000523709600007&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0098-3004
IngestDate Thu Oct 02 06:35:39 EDT 2025
Tue Nov 18 21:07:20 EST 2025
Sat Nov 29 07:22:32 EST 2025
Fri Feb 23 02:48:42 EST 2024
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Keywords Model order reduction
Randomized singular value decomposition
Krylov subspaces
Tikhonov regularization
Regularization parameter
Language English
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-a404t-54d77c4e14b29f80cf6d71c63ec9e2124bfa49b9155e5f21841ed6b4aa85fe973
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
OpenAccessLink https://www.sciencedirect.com/science/article/pii/S0098300418306952
PQID 2400457481
PQPubID 24069
ParticipantIDs proquest_miscellaneous_2400457481
crossref_primary_10_1016_j_cageo_2020_104427
crossref_citationtrail_10_1016_j_cageo_2020_104427
elsevier_sciencedirect_doi_10_1016_j_cageo_2020_104427
PublicationCentury 2000
PublicationDate April 2020
2020-04-00
20200401
PublicationDateYYYYMMDD 2020-04-01
PublicationDate_xml – month: 04
  year: 2020
  text: April 2020
PublicationDecade 2020
PublicationTitle Computers & geosciences
PublicationYear 2020
Publisher Elsevier Ltd
Publisher_xml – name: Elsevier Ltd
References Kilmer, O’Leary (b21) 2001; 22
Bauer, Lukas (b3) 2011; 81
Neupauer, Borchers (b30) 2001; 27
Leonov (b23) 1991; 44
Lukas (b25) 2008; 24
Morozov (b29) 1984
Hansen (b14) 1992; 34
Mallows (b26) 1973; 15
Mejer Hansen, Mosegaard (b27) 2008; 34
Stoer, Bulirsch (b32) 1983
Wahba (b37) 1990
Halko, Martinsson, Tropp (b12) 2011; 53
Vogel (b36) 2002
Hansen, Saxild-Hansen (b19) 2012; 236
Vatankhah, Renaut, Ardestani (b35) 2018; 83
Leonov (b22) 1978; 19
Hansen (b15) 1998
Calvetti, Golub, Reichel (b4) 1999; 39
Hansen (b13) 1987; 27
Xiang, Zou (b40) 2015; 29
Bakushinskii (b2) 1981; 24
Golub, von Matt (b9) 1997
Hutchinson (b20) 1990; 19
Reginska (b31) 1996; 17
Xiang, Zou (b39) 2013; 29
Bai, Fahey, Golub (b1) 1996; 74
Zunino, Mosegaard (b41) 2019; 122
Golub, Meurant (b10) 2009
Chung, Nagy, O’Leary (b5) 2008; 28
Varah (b34) 1983; 4
Hansen (b16) 2010
Hansen (b17) 2013; 20
Morozov (b28) 1984
Hansen, O’Leary (b18) 1993; 14
Lukas (b24) 2006; 22
Golub, von Matt (b8) 1995; 6
Golub, Heath, Wahba (b7) 1979; 21
Wei, Xie, Zhang (b38) 2016; 37
Golub, Van Loan (b11) 2013
Ubaru, Chen, Saad (b33) 2017; 38
Engl, Hanke, Neubauer (b6) 1996
Hansen (10.1016/j.cageo.2020.104427_b17) 2013; 20
Leonov (10.1016/j.cageo.2020.104427_b22) 1978; 19
Ubaru (10.1016/j.cageo.2020.104427_b33) 2017; 38
Hansen (10.1016/j.cageo.2020.104427_b13) 1987; 27
Bakushinskii (10.1016/j.cageo.2020.104427_b2) 1981; 24
Lukas (10.1016/j.cageo.2020.104427_b25) 2008; 24
Golub (10.1016/j.cageo.2020.104427_b11) 2013
Hansen (10.1016/j.cageo.2020.104427_b18) 1993; 14
Chung (10.1016/j.cageo.2020.104427_b5) 2008; 28
Hansen (10.1016/j.cageo.2020.104427_b19) 2012; 236
Neupauer (10.1016/j.cageo.2020.104427_b30) 2001; 27
Lukas (10.1016/j.cageo.2020.104427_b24) 2006; 22
Engl (10.1016/j.cageo.2020.104427_b6) 1996
Morozov (10.1016/j.cageo.2020.104427_b28) 1984
Bauer (10.1016/j.cageo.2020.104427_b3) 2011; 81
Wei (10.1016/j.cageo.2020.104427_b38) 2016; 37
Varah (10.1016/j.cageo.2020.104427_b34) 1983; 4
Hutchinson (10.1016/j.cageo.2020.104427_b20) 1990; 19
Vatankhah (10.1016/j.cageo.2020.104427_b35) 2018; 83
Xiang (10.1016/j.cageo.2020.104427_b40) 2015; 29
Calvetti (10.1016/j.cageo.2020.104427_b4) 1999; 39
Hansen (10.1016/j.cageo.2020.104427_b14) 1992; 34
Leonov (10.1016/j.cageo.2020.104427_b23) 1991; 44
Hansen (10.1016/j.cageo.2020.104427_b15) 1998
Zunino (10.1016/j.cageo.2020.104427_b41) 2019; 122
Wahba (10.1016/j.cageo.2020.104427_b37) 1990
Bai (10.1016/j.cageo.2020.104427_b1) 1996; 74
Golub (10.1016/j.cageo.2020.104427_b8) 1995; 6
Halko (10.1016/j.cageo.2020.104427_b12) 2011; 53
Mejer Hansen (10.1016/j.cageo.2020.104427_b27) 2008; 34
Kilmer (10.1016/j.cageo.2020.104427_b21) 2001; 22
Hansen (10.1016/j.cageo.2020.104427_b16) 2010
Mallows (10.1016/j.cageo.2020.104427_b26) 1973; 15
Reginska (10.1016/j.cageo.2020.104427_b31) 1996; 17
Golub (10.1016/j.cageo.2020.104427_b10) 2009
Golub (10.1016/j.cageo.2020.104427_b9) 1997
Golub (10.1016/j.cageo.2020.104427_b7) 1979; 21
Xiang (10.1016/j.cageo.2020.104427_b39) 2013; 29
Morozov (10.1016/j.cageo.2020.104427_b29) 1984
Vogel (10.1016/j.cageo.2020.104427_b36) 2002
Stoer (10.1016/j.cageo.2020.104427_b32) 1983
References_xml – volume: 21
  start-page: 215
  year: 1979
  end-page: 223
  ident: b7
  article-title: Generalized cross-validation as a method for choosing a good ridge parameter
  publication-title: Technometrics
– volume: 236
  start-page: 2167
  year: 2012
  end-page: 2178
  ident: b19
  article-title: AIR Tools — A MATLAB package of algebraic iterative reconstruction methods
  publication-title: J. Comput. Appl. Math.
– volume: 38
  start-page: 1075
  year: 2017
  end-page: 1099
  ident: b33
  article-title: Fast estimation of $tr(f(a))$ via stochastic lanczos quadrature
  publication-title: SIAM J. Matrix Anal. Appl.
– year: 2009
  ident: b10
  article-title: Matrices, Moments and Quadrature with Applications
– volume: 19
  start-page: 537
  year: 1978
  end-page: 540
  ident: b22
  article-title: On the choice of regularization parameters by means of the quasi-optimality and ratio criteria
  publication-title: Svoiet Math. Dokl.
– volume: 20
  start-page: 250
  year: 2013
  end-page: 258
  ident: b17
  article-title: Oblique projections and standard-form transformations for discrete inverse problems
  publication-title: Numer. Linear Algebra Appl.
– volume: 29
  year: 2013
  ident: b39
  article-title: Regularization with randomized SVD for large-scale discrete inverse problems
  publication-title: Inverse Problems
– year: 2013
  ident: b11
  publication-title: Matrix Computations
– volume: 19
  start-page: 433
  year: 1990
  end-page: 450
  ident: b20
  article-title: A stochastic estimator for the trace of the influence matrix for Laplacian smoothing splines
  publication-title: Comm. Statist. Simulation Comput.
– volume: 81
  start-page: 1795
  year: 2011
  end-page: 1841
  ident: b3
  article-title: Comparing parameter choice methods for regularization of ill-posed problems
  publication-title: Math. Comput. Simulation
– start-page: 3
  year: 1997
  end-page: 26
  ident: b9
  article-title: Tikhonov regularization for large scale problems
  publication-title: Workshop on Scientific Computing
– volume: 24
  start-page: 034006
  year: 2008
  ident: b25
  article-title: Strong robust generalized cross-validation for choosing the regularization parameter
  publication-title: Inverse Problems
– volume: 22
  start-page: 1204
  year: 2001
  end-page: 1221
  ident: b21
  article-title: Choosing regularization parameters in iterative methods for ill-posed problems
  publication-title: SIAM J. Matrix Anal. Appl.
– volume: 15
  start-page: 661
  year: 1973
  end-page: 675
  ident: b26
  article-title: Some comments on Cp
  publication-title: Technometrics
– year: 1990
  ident: b37
  article-title: Spline Models for Observational Data
– volume: 27
  start-page: 757
  year: 2001
  end-page: 762
  ident: b30
  article-title: A MATLAB implementation of the minimum relative entropy method for linear inverse problems
  publication-title: Comput. Geosci.
– year: 1983
  ident: b32
  article-title: Introduction to Numerical Analysis
– year: 2010
  ident: b16
  article-title: Discrete Inverse Problems: Insight and Algorithms
– volume: 34
  start-page: 561
  year: 1992
  end-page: 580
  ident: b14
  article-title: Analysis of discrete ill-posed problems by means of the L-curve
  publication-title: SIAM Rev.
– volume: 34
  start-page: 53
  year: 2008
  end-page: 76
  ident: b27
  article-title: VISIM: Sequential simulation for linear inverse problems
  publication-title: Comput. Geosci.
– volume: 37
  start-page: 649
  year: 2016
  end-page: 675
  ident: b38
  article-title: Tikhonov regularization and randomized GSVD
  publication-title: SIAM J. Matrix Anal. Appl.
– volume: 4
  start-page: 164
  year: 1983
  end-page: 176
  ident: b34
  article-title: Pitfalls in the numerical solution of linear ill-posed problems
  publication-title: SIAM J. Stat. Sci. Comput.
– volume: 28
  start-page: 149
  year: 2008
  end-page: 167
  ident: b5
  article-title: A weighted GCV method for lanczos hybrid regularization
  publication-title: Electron. Trans. Numer. Anal.
– volume: 74
  start-page: 71
  year: 1996
  end-page: 89
  ident: b1
  article-title: Some large-scale matrix computation problems
  publication-title: J. Comput. Appl. Math.
– year: 1996
  ident: b6
  article-title: Regularization of Inverse Problems
– year: 1984
  ident: b29
  article-title: Methods for Solving Incorrectly Posed Problems
– volume: 39
  start-page: 603
  year: 1999
  end-page: 619
  ident: b4
  article-title: Estimation of the L-curve via lanczos bidiagonalization
  publication-title: BIT Numer. Math.
– volume: 22
  start-page: 1883
  year: 2006
  ident: b24
  article-title: Robust generalized cross-validation for choosing the regularization parameter
  publication-title: Inverse Problems
– volume: 27
  start-page: 534
  year: 1987
  end-page: 553
  ident: b13
  article-title: The truncated SVD as a method for regularization
  publication-title: BIT
– volume: 122
  start-page: 77
  year: 2019
  end-page: 86
  ident: b41
  article-title: An efficient method to solve large linearizable inverse problems under Gaussian and separability assumptions
  publication-title: Comput. Geosci.
– volume: 24
  start-page: 181
  year: 1981
  end-page: 182
  ident: b2
  article-title: Remarks on choosing a regularization parameter usnig the quasi-optimality and ratio criterion
  publication-title: USSR Comput. Math. Math. Phys.
– year: 2002
  ident: b36
  article-title: Computational Methods for Inverse Problems
– volume: 29
  year: 2015
  ident: b40
  article-title: Randomized algorithms for large-scale inverse problems with general form Tikhonov regularization
  publication-title: Inverse Problems
– year: 1984
  ident: b28
  article-title: Methods for Solving Incorrectly Posed Problems
– volume: 14
  start-page: 1487
  year: 1993
  end-page: 1503
  ident: b18
  article-title: The use of the L-curve in the regularization of discrete ill-posed problems
  publication-title: SIAM J. Sci. Comput.
– volume: 83
  start-page: 25
  year: 2018
  end-page: 34
  ident: b35
  article-title: A fast algorithm for regularized focused 3D inversion of gravity data using randomized singular-value decomposition
  publication-title: Geophysics
– volume: 53
  start-page: 217
  year: 2011
  end-page: 288
  ident: b12
  article-title: Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions
  publication-title: SIAM Rev.
– year: 1998
  ident: b15
  article-title: Rank-Deficient and Discrete Ill-Posed Problems
– volume: 44
  start-page: 711
  year: 1991
  end-page: 716
  ident: b23
  article-title: On the accuracy of tikhonov regularizing algorithms and quasioptimal selection of a regularization parameter
  publication-title: Soviet Math. Dokl.
– volume: 6
  start-page: 1
  year: 1995
  end-page: 34
  ident: b8
  article-title: Generalized cross-validation for large scale problems
  publication-title: J. Comput. Graph. Statist.
– volume: 17
  year: 1996
  ident: b31
  article-title: A regularization parameter in discrete ill-posed problems
  publication-title: SIAM J. Sci. Comput.
– year: 2009
  ident: 10.1016/j.cageo.2020.104427_b10
– volume: 44
  start-page: 711
  year: 1991
  ident: 10.1016/j.cageo.2020.104427_b23
  article-title: On the accuracy of tikhonov regularizing algorithms and quasioptimal selection of a regularization parameter
  publication-title: Soviet Math. Dokl.
– year: 1983
  ident: 10.1016/j.cageo.2020.104427_b32
– start-page: 3
  year: 1997
  ident: 10.1016/j.cageo.2020.104427_b9
  article-title: Tikhonov regularization for large scale problems
– volume: 53
  start-page: 217
  issue: 2
  year: 2011
  ident: 10.1016/j.cageo.2020.104427_b12
  article-title: Finding structure with randomness: Probabilistic algorithms for constructing approximate matrix decompositions
  publication-title: SIAM Rev.
  doi: 10.1137/090771806
– volume: 24
  start-page: 181
  year: 1981
  ident: 10.1016/j.cageo.2020.104427_b2
  article-title: Remarks on choosing a regularization parameter usnig the quasi-optimality and ratio criterion
  publication-title: USSR Comput. Math. Math. Phys.
  doi: 10.1016/0041-5553(84)90253-2
– volume: 34
  start-page: 53
  year: 2008
  ident: 10.1016/j.cageo.2020.104427_b27
  article-title: VISIM: Sequential simulation for linear inverse problems
  publication-title: Comput. Geosci.
  doi: 10.1016/j.cageo.2007.02.003
– volume: 21
  start-page: 215
  issue: 2
  year: 1979
  ident: 10.1016/j.cageo.2020.104427_b7
  article-title: Generalized cross-validation as a method for choosing a good ridge parameter
  publication-title: Technometrics
  doi: 10.1080/00401706.1979.10489751
– year: 2010
  ident: 10.1016/j.cageo.2020.104427_b16
– volume: 14
  start-page: 1487
  issue: 6
  year: 1993
  ident: 10.1016/j.cageo.2020.104427_b18
  article-title: The use of the L-curve in the regularization of discrete ill-posed problems
  publication-title: SIAM J. Sci. Comput.
  doi: 10.1137/0914086
– volume: 83
  start-page: 25
  issue: 4
  year: 2018
  ident: 10.1016/j.cageo.2020.104427_b35
  article-title: A fast algorithm for regularized focused 3D inversion of gravity data using randomized singular-value decomposition
  publication-title: Geophysics
  doi: 10.1190/geo2017-0386.1
– volume: 39
  start-page: 603
  issue: 4
  year: 1999
  ident: 10.1016/j.cageo.2020.104427_b4
  article-title: Estimation of the L-curve via lanczos bidiagonalization
  publication-title: BIT Numer. Math.
  doi: 10.1023/A:1022383005969
– year: 2002
  ident: 10.1016/j.cageo.2020.104427_b36
– volume: 4
  start-page: 164
  issue: 2
  year: 1983
  ident: 10.1016/j.cageo.2020.104427_b34
  article-title: Pitfalls in the numerical solution of linear ill-posed problems
  publication-title: SIAM J. Stat. Sci. Comput.
  doi: 10.1137/0904012
– volume: 22
  start-page: 1204
  issue: 4
  year: 2001
  ident: 10.1016/j.cageo.2020.104427_b21
  article-title: Choosing regularization parameters in iterative methods for ill-posed problems
  publication-title: SIAM J. Matrix Anal. Appl.
  doi: 10.1137/S0895479899345960
– volume: 29
  year: 2015
  ident: 10.1016/j.cageo.2020.104427_b40
  article-title: Randomized algorithms for large-scale inverse problems with general form Tikhonov regularization
  publication-title: Inverse Problems
– year: 1984
  ident: 10.1016/j.cageo.2020.104427_b28
– year: 1990
  ident: 10.1016/j.cageo.2020.104427_b37
– year: 1998
  ident: 10.1016/j.cageo.2020.104427_b15
– volume: 24
  start-page: 034006
  issue: 3
  year: 2008
  ident: 10.1016/j.cageo.2020.104427_b25
  article-title: Strong robust generalized cross-validation for choosing the regularization parameter
  publication-title: Inverse Problems
  doi: 10.1088/0266-5611/24/3/034006
– volume: 28
  start-page: 149
  year: 2008
  ident: 10.1016/j.cageo.2020.104427_b5
  article-title: A weighted GCV method for lanczos hybrid regularization
  publication-title: Electron. Trans. Numer. Anal.
– year: 1984
  ident: 10.1016/j.cageo.2020.104427_b29
– volume: 27
  start-page: 534
  year: 1987
  ident: 10.1016/j.cageo.2020.104427_b13
  article-title: The truncated SVD as a method for regularization
  publication-title: BIT
  doi: 10.1007/BF01937276
– volume: 122
  start-page: 77
  year: 2019
  ident: 10.1016/j.cageo.2020.104427_b41
  article-title: An efficient method to solve large linearizable inverse problems under Gaussian and separability assumptions
  publication-title: Comput. Geosci.
  doi: 10.1016/j.cageo.2018.09.005
– volume: 81
  start-page: 1795
  issue: 9
  year: 2011
  ident: 10.1016/j.cageo.2020.104427_b3
  article-title: Comparing parameter choice methods for regularization of ill-posed problems
  publication-title: Math. Comput. Simulation
  doi: 10.1016/j.matcom.2011.01.016
– volume: 22
  start-page: 1883
  issue: 5
  year: 2006
  ident: 10.1016/j.cageo.2020.104427_b24
  article-title: Robust generalized cross-validation for choosing the regularization parameter
  publication-title: Inverse Problems
  doi: 10.1088/0266-5611/22/5/021
– volume: 29
  issue: 8
  year: 2013
  ident: 10.1016/j.cageo.2020.104427_b39
  article-title: Regularization with randomized SVD for large-scale discrete inverse problems
  publication-title: Inverse Problems
  doi: 10.1088/0266-5611/29/8/085008
– year: 1996
  ident: 10.1016/j.cageo.2020.104427_b6
– volume: 6
  start-page: 1
  year: 1995
  ident: 10.1016/j.cageo.2020.104427_b8
  article-title: Generalized cross-validation for large scale problems
  publication-title: J. Comput. Graph. Statist.
  doi: 10.1080/10618600.1997.10474725
– volume: 74
  start-page: 71
  year: 1996
  ident: 10.1016/j.cageo.2020.104427_b1
  article-title: Some large-scale matrix computation problems
  publication-title: J. Comput. Appl. Math.
  doi: 10.1016/0377-0427(96)00018-0
– volume: 15
  start-page: 661
  year: 1973
  ident: 10.1016/j.cageo.2020.104427_b26
  article-title: Some comments on Cp
  publication-title: Technometrics
– volume: 19
  start-page: 433
  year: 1990
  ident: 10.1016/j.cageo.2020.104427_b20
  article-title: A stochastic estimator for the trace of the influence matrix for Laplacian smoothing splines
  publication-title: Comm. Statist. Simulation Comput.
  doi: 10.1080/03610919008812866
– year: 2013
  ident: 10.1016/j.cageo.2020.104427_b11
– volume: 19
  start-page: 537
  year: 1978
  ident: 10.1016/j.cageo.2020.104427_b22
  article-title: On the choice of regularization parameters by means of the quasi-optimality and ratio criteria
  publication-title: Svoiet Math. Dokl.
– volume: 38
  start-page: 1075
  issue: 4
  year: 2017
  ident: 10.1016/j.cageo.2020.104427_b33
  article-title: Fast estimation of $tr(f(a))$ via stochastic lanczos quadrature
  publication-title: SIAM J. Matrix Anal. Appl.
  doi: 10.1137/16M1104974
– volume: 37
  start-page: 649
  year: 2016
  ident: 10.1016/j.cageo.2020.104427_b38
  article-title: Tikhonov regularization and randomized GSVD
  publication-title: SIAM J. Matrix Anal. Appl.
  doi: 10.1137/15M1030200
– volume: 34
  start-page: 561
  year: 1992
  ident: 10.1016/j.cageo.2020.104427_b14
  article-title: Analysis of discrete ill-posed problems by means of the L-curve
  publication-title: SIAM Rev.
  doi: 10.1137/1034115
– volume: 27
  start-page: 757
  year: 2001
  ident: 10.1016/j.cageo.2020.104427_b30
  article-title: A MATLAB implementation of the minimum relative entropy method for linear inverse problems
  publication-title: Comput. Geosci.
  doi: 10.1016/S0098-3004(01)00009-7
– volume: 236
  start-page: 2167
  issue: 8
  year: 2012
  ident: 10.1016/j.cageo.2020.104427_b19
  article-title: AIR Tools — A MATLAB package of algebraic iterative reconstruction methods
  publication-title: J. Comput. Appl. Math.
  doi: 10.1016/j.cam.2011.09.039
– volume: 17
  year: 1996
  ident: 10.1016/j.cageo.2020.104427_b31
  article-title: A regularization parameter in discrete ill-posed problems
  publication-title: SIAM J. Sci. Comput.
  doi: 10.1137/S1064827593252672
– volume: 20
  start-page: 250
  year: 2013
  ident: 10.1016/j.cageo.2020.104427_b17
  article-title: Oblique projections and standard-form transformations for discrete inverse problems
  publication-title: Numer. Linear Algebra Appl.
  doi: 10.1002/nla.802
SSID ssj0002285
Score 2.3048596
Snippet In this work we address regularization parameter estimation for ill-posed linear inverse problems with an ℓ2 penalty. Regularization parameter selection is of...
SourceID proquest
crossref
elsevier
SourceType Aggregation Database
Enrichment Source
Index Database
Publisher
StartPage 104427
SubjectTerms algorithms
computers
geostatistics
Krylov subspaces
Model order reduction
model validation
Randomized singular value decomposition
Regularization parameter
spatial data
Tikhonov regularization
Title Comparing RSVD and Krylov methods for linear inverse problems
URI https://dx.doi.org/10.1016/j.cageo.2020.104427
https://www.proquest.com/docview/2400457481
Volume 137
WOSCitedRecordID wos000523709600007&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVESC
  databaseName: Elsevier SD Freedom Collection Journals 2021
  customDbUrl:
  eissn: 1873-7803
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0002285
  issn: 0098-3004
  databaseCode: AIEXJ
  dateStart: 19950101
  isFulltext: true
  titleUrlDefault: https://www.sciencedirect.com
  providerName: Elsevier
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1Lb9QwELZgCxIXxFO0PGQkeiqREseO42PV7vKKFiTSam9W7HWqFpRuN7ul_Hs8fuyWV1UOXKLIayer-Ubjmcx4PoReCaEa2pJpoohmNkApddK0bZMQZhfwNqWZ41g6rPh4XE4m4lNg1OsdnQDvuvLiQsz-K9R2zIINR2f_Ae7VQ-2Avbeg26uF3V6vBfyeZxaEqrrPh_suOfBhbsPy88AW7Row7IB3CfWnHZRlwGEpxyvTX_ZVI-FD79TjyIS2l-uiw2p5_MVbrXHsrR8ORFXGLL_5n2qwIkEFw9cFkl4qSgkWU5QJdOX6yWLmfGcG2WFKePJHO-w_CZzYGPvIHbEkaZi-3nZiqn38UY4OqkrWw0m9nY9mZwlQgkHqfDvf9_DcRBuEM1EO0Mbuu-Hk_WqrJaRksSkq_MnYVsoV8P327r-5Hr9sws6zqO-huyEkwLseyvvohukeoNtvHOXy94doDSgGQLEFFHtAcQAUW0CxBxQHQHEE9BE6GA3rvbdJIL1IGprSRcLolHNNTUYVEW2Z6raY8kwXudHCWD-DqrahQkFbf8NaCNAzMy0UbZoS6gZ5_hgNutPOPEHYDhSqaEWW55pOTS6YICxX1sVjimmiNhGJ8pA6dIQHYpKvMpb-nUgnRAlClF6Im-j1atHMN0S5enoRBS2DlnpfTVpVuXrhywiLtBYP0lhNZ06XvYSqZ8o4LbOta8x5iu6s9foZGizmS_Mc3dLni-N-_iIo1A-hTHVj
linkProvider Elsevier
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Comparing+RSVD+and+Krylov+methods+for+linear+inverse+problems&rft.jtitle=Computers+%26+geosciences&rft.au=Luiken%2C+Nick&rft.au=van+Leeuwen%2C+Tristan&rft.date=2020-04-01&rft.issn=0098-3004&rft.volume=137+p.104427-&rft_id=info:doi/10.1016%2Fj.cageo.2020.104427&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0098-3004&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0098-3004&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0098-3004&client=summon