Norris, BoyanaSood, Kanika2019-09-182019-09-182019-09-18https://hdl.handle.net/1794/24931Scientific and engineering applications are dominated by linear algebra and depend on scalable solutions of sparse linear systems. For large problems, preconditioned iterative methods are a popular choice. High-performance numerical libraries offer a variety of preconditioned Newton-Krylov methods for solving sparse problems. However, the selection of a well-performing Krylov method remains to be the user’s responsibility. This research presents the technique for choosing well-performing parallel sparse linear solver methods, based on the problem characteristics and the amount of communication involved in the Krylov methodsen-USAll Rights Reserved.machine learningnumerical libraryPETScsolver selectionsparse linear systemsElectronic Thesis or Dissertation