BEGIN:VCALENDAR
PRODID:-//Microsoft Corporation//Outlook MIMEDIR//EN
VERSION:2.0
BEGIN:VEVENT
DTSTART:20151117T231500Z
DTEND:20151118T010000Z
LOCATION:Level 4 - Lobby
DESCRIPTION;ENCODING=QUOTED-PRINTABLE:ABSTRACT: Solving large, sparse linear systems  efficiently is a=0Achallenging problem in scientific computing. Taxonomies=0Aand high-performance numerical linear algebra solutions=0Ahelp to translate algorithms to =0Asoftware.  However, accessible, comprehensive, and usable tools=0Afor high quality code production are not available. To address this challenge,=0Awe present an extensible methodology for classifying iterative=0Aalgorithms for solving sparse linear systems.  =0ALighthouse is the first framework that offers an organized=0Ataxonomy of software components for linear algebra that enables=0Afunctionality and performance-based search and generates code templates=0Aand optimized low-level kernels.  It enables the=0Aselection of a solution method that is likely to converge and perform=0Awell.  We describe the integration of PETSc and=0ATrilinos iterative solvers into Lighthouse. =0AWe present a comparative analysis of solver classification=0Aresults for a varied set of input problems and machine learning methods=0Aachieving up to 93% accuracy in identifying the best-performing linear=0Asolution methods.
SUMMARY:Automating Sparse Linear Solver Selection with Lighthouse
PRIORITY:3
END:VEVENT
END:VCALENDAR