Reusing and Updating Preconditioners for Sequences of Matrices

dc.contributor.authorGrim-McNally, Arielle Katherineen
dc.contributor.committeechairde Sturler, Ericen
dc.contributor.committeememberGugercin, Serkanen
dc.contributor.committeememberChung, Julianneen
dc.contributor.departmentMathematicsen
dc.date.accessioned2015-06-17T08:00:35Zen
dc.date.available2015-06-17T08:00:35Zen
dc.date.issued2015-06-15en
dc.description.abstractFor sequences of related linear systems, the computation of a preconditioner for every system can be expensive. Often a fixed preconditioner is used, but this may not be effective as the matrix changes. This research examines the benefits of both reusing and recycling preconditioners, with special focus on ILUTP and factorized sparse approximate inverses and proposes an update that we refer to as a sparse approximate map or SAM update. Analysis of the residual and eigenvalues of the map will be provided. Applications include the Quantum Monte Carlo method, model reduction, oscillatory hydraulic tomography, diffuse optical tomography, and Helmholtz-type problems.en
dc.description.degreeMaster of Scienceen
dc.format.mediumETDen
dc.identifier.othervt_gsexam:5645en
dc.identifier.urihttp://hdl.handle.net/10919/52945en
dc.publisherVirginia Techen
dc.rightsIn Copyrighten
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/en
dc.subjectRecycling Preconditionersen
dc.subjectPreconditionersen
dc.subjectSparse Approximate Mapen
dc.subjectIncomplete LU Decompositionen
dc.subjectSparse Approximate Inverseen
dc.subjectFactorized Sparse Approximate Inverseen
dc.subjectKrylov Methodsen
dc.titleReusing and Updating Preconditioners for Sequences of Matricesen
dc.typeThesisen
thesis.degree.disciplineMathematicsen
thesis.degree.grantorVirginia Polytechnic Institute and State Universityen
thesis.degree.levelmastersen
thesis.degree.nameMaster of Scienceen

Files

Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Grim-McNally_AK_T_2015.pdf
Size:
1.02 MB
Format:
Adobe Portable Document Format

Collections