A combinatorial scheme for developing efficient composite solvers

Sanjukta Bhowmick; Padma Raghavan; Keita Teranishi

doi:10.1007/3-540-46080-2_34

A combinatorial scheme for developing efficient composite solvers

Sanjukta Bhowmick, Padma Raghavan, Keita Teranishi

Computer Science

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

9 Scopus citations

Abstract

Many fundamental problems in scientific computing have more than one solution method. It is not uncommon for alternative solution methods to represent different tradeoffs between solution cost and reliability. Furthermore, the performance of a solution method often depends on the numerical properties of the problem instance and thus can vary dramatically across application domains. In such situations, it is natural to consider the construction of a multi-method composite solver to potentially improve both the average performance and reliability. In this paper, we provide a combinatorial framework for developing such composite solvers. We provide analytical results for obtaining an optimal composite from a set of methods with normalized measures of performance and reliability. Our empirical results demonstrate the effectiveness of such optimal composites for solving large, sparse linear systems of equations.

Original language	English (US)
Title of host publication	Computational Science, ICCS 2002 - International Conference, Proceedings
Publisher	Springer Verlag
Pages	325-334
Number of pages	10
Edition	PART 2
ISBN (Print)	354043593X, 9783540435938
DOIs	https://doi.org/10.1007/3-540-46080-2_34
State	Published - 2002
Event	International Conference on Computational Science, ICCS 2002 - Amsterdam, Netherlands Duration: Apr 21 2002 → Apr 24 2002

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Number	PART 2
Volume	2330 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	International Conference on Computational Science, ICCS 2002
Country/Territory	Netherlands
City	Amsterdam
Period	4/21/02 → 4/24/02

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/3-540-46080-2_34

Cite this

Bhowmick, S., Raghavan, P., & Teranishi, K. (2002). A combinatorial scheme for developing efficient composite solvers. In Computational Science, ICCS 2002 - International Conference, Proceedings (PART 2 ed., pp. 325-334). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2330 LNCS, No. PART 2). Springer Verlag. https://doi.org/10.1007/3-540-46080-2_34

A combinatorial scheme for developing efficient composite solvers. / Bhowmick, Sanjukta; Raghavan, Padma; Teranishi, Keita.
Computational Science, ICCS 2002 - International Conference, Proceedings. PART 2. ed. Springer Verlag, 2002. p. 325-334 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 2330 LNCS, No. PART 2).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Bhowmick, S, Raghavan, P & Teranishi, K 2002, A combinatorial scheme for developing efficient composite solvers. in Computational Science, ICCS 2002 - International Conference, Proceedings. PART 2 edn, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), no. PART 2, vol. 2330 LNCS, Springer Verlag, pp. 325-334, International Conference on Computational Science, ICCS 2002, Amsterdam, Netherlands, 4/21/02. https://doi.org/10.1007/3-540-46080-2_34

Bhowmick S, Raghavan P, Teranishi K. A combinatorial scheme for developing efficient composite solvers. In Computational Science, ICCS 2002 - International Conference, Proceedings. PART 2 ed. Springer Verlag. 2002. p. 325-334. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2). doi: 10.1007/3-540-46080-2_34

Bhowmick, Sanjukta ; Raghavan, Padma ; Teranishi, Keita. / A combinatorial scheme for developing efficient composite solvers. Computational Science, ICCS 2002 - International Conference, Proceedings. PART 2. ed. Springer Verlag, 2002. pp. 325-334 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); PART 2).

@inproceedings{6bd5759ccf1c49109a1c5ab2279d6825,

title = "A combinatorial scheme for developing efficient composite solvers",

abstract = "Many fundamental problems in scientific computing have more than one solution method. It is not uncommon for alternative solution methods to represent different tradeoffs between solution cost and reliability. Furthermore, the performance of a solution method often depends on the numerical properties of the problem instance and thus can vary dramatically across application domains. In such situations, it is natural to consider the construction of a multi-method composite solver to potentially improve both the average performance and reliability. In this paper, we provide a combinatorial framework for developing such composite solvers. We provide analytical results for obtaining an optimal composite from a set of methods with normalized measures of performance and reliability. Our empirical results demonstrate the effectiveness of such optimal composites for solving large, sparse linear systems of equations.",

author = "Sanjukta Bhowmick and Padma Raghavan and Keita Teranishi",

year = "2002",

doi = "10.1007/3-540-46080-2_34",

language = "English (US)",

isbn = "354043593X",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer Verlag",

number = "PART 2",

pages = "325--334",

booktitle = "Computational Science, ICCS 2002 - International Conference, Proceedings",

edition = "PART 2",

note = "International Conference on Computational Science, ICCS 2002 ; Conference date: 21-04-2002 Through 24-04-2002",

}

TY - GEN

T1 - A combinatorial scheme for developing efficient composite solvers

AU - Bhowmick, Sanjukta

AU - Raghavan, Padma

AU - Teranishi, Keita

PY - 2002

Y1 - 2002

N2 - Many fundamental problems in scientific computing have more than one solution method. It is not uncommon for alternative solution methods to represent different tradeoffs between solution cost and reliability. Furthermore, the performance of a solution method often depends on the numerical properties of the problem instance and thus can vary dramatically across application domains. In such situations, it is natural to consider the construction of a multi-method composite solver to potentially improve both the average performance and reliability. In this paper, we provide a combinatorial framework for developing such composite solvers. We provide analytical results for obtaining an optimal composite from a set of methods with normalized measures of performance and reliability. Our empirical results demonstrate the effectiveness of such optimal composites for solving large, sparse linear systems of equations.

AB - Many fundamental problems in scientific computing have more than one solution method. It is not uncommon for alternative solution methods to represent different tradeoffs between solution cost and reliability. Furthermore, the performance of a solution method often depends on the numerical properties of the problem instance and thus can vary dramatically across application domains. In such situations, it is natural to consider the construction of a multi-method composite solver to potentially improve both the average performance and reliability. In this paper, we provide a combinatorial framework for developing such composite solvers. We provide analytical results for obtaining an optimal composite from a set of methods with normalized measures of performance and reliability. Our empirical results demonstrate the effectiveness of such optimal composites for solving large, sparse linear systems of equations.

UR - http://www.scopus.com/inward/record.url?scp=23044532053&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=23044532053&partnerID=8YFLogxK

U2 - 10.1007/3-540-46080-2_34

DO - 10.1007/3-540-46080-2_34

M3 - Conference contribution

AN - SCOPUS:23044532053

SN - 354043593X

SN - 9783540435938

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 325

EP - 334

BT - Computational Science, ICCS 2002 - International Conference, Proceedings

PB - Springer Verlag

T2 - International Conference on Computational Science, ICCS 2002

Y2 - 21 April 2002 through 24 April 2002

ER -

A combinatorial scheme for developing efficient composite solvers

Abstract

Publication series

Conference

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this