Show simple item record

dc.contributor.advisorQasem, Apan
dc.contributor.authorSaha, Biplab Kumar
dc.date.accessioned2016-11-07T21:59:11Z
dc.date.available2016-11-07T21:59:11Z
dc.date.created2016-08
dc.date.issued2016-07-27
dc.date.submittedAugust 2016
dc.identifier.urihttps://digital.library.txstate.edu/handle/10877/6343
dc.description.abstractRecent interest in machine learning-based methods have produced many sophisticated models for performance modeling and optimi:,ation. These models tend to be sensitive to architectural parameters and are most effective when trained on the target platform. Training of these models; however; is a fairly involved process and requires knowledge of statistics and machine learning that the end-users of such models may not possess. This paper presents a framework for automatically generating machine learning-based performance models. Leveraging existing open-source software; we provide a tool-chain that provides automated mechanisms for sample generation; dynamic feature extraction; feature selection; data labeling; validation and model selection. We describe the design of the framework and demonstrate its effectiveness by developing a learning heuristic for register allocation of GPU kernels. The results show the newly created models are accurate and can predict register caps that lead to substantial improvements in execution time without incurring a penalty in power consumption.
dc.formatText
dc.format.extent71 pages
dc.format.medium1 file (.pdf)
dc.language.isoen_US
dc.subjectHigh performance computing
dc.subjectMachine learning
dc.subject.lcshHigh performance computing
dc.subject.lcshMachine learning
dc.titleTowards a Framework for Automating the Workflow for Building Machine Learning Based Performance Tuning
txstate.documenttypeThesis
dc.contributor.committeeMemberEkstrand, Michael
dc.contributor.committeeMemberMetsis, Vangelis
thesis.degree.departmentComputer Scienceen_US
thesis.degree.disciplineComputer Scienceen_US
thesis.degree.grantorTexas State Universityen_US
thesis.degree.levelMastersen_US
thesis.degree.nameMaster of Scienceen_US
txstate.departmentComputer Science


Download

Thumbnail

This item appears in the following Collection(s)

Show simple item record