Electrical Engineering
      and Computer Sciences

Electrical Engineering and Computer Sciences

COLLEGE OF ENGINEERING

UC Berkeley

Fast Support Vector Machine Training and Classification on Graphics Processors

Bryan Christopher Catanzaro, Narayanan Sundaram and Kurt Keutzer

EECS Department
University of California, Berkeley
Technical Report No. UCB/EECS-2008-11
February 8, 2008

http://www.eecs.berkeley.edu/Pubs/TechRpts/2008/EECS-2008-11.pdf

Recent developments in programmable, highly parallel Graphics Processing Units (GPUs) have enabled high performance implementations of machine learning algorithms. We describe a solver for Support Vector Machine training, using Platt's Sequential Minimal Optimization algorithm, which achieves speedups of 5-32x over LibSVM running on a high-end traditional processor. We also present a system for SVM classification which achieves speedups of 120-150x over LibSVM.


BibTeX citation:

@techreport{Catanzaro:EECS-2008-11,
    Author = {Catanzaro, Bryan Christopher and Sundaram, Narayanan and Keutzer, Kurt},
    Title = {Fast Support Vector Machine Training and Classification on Graphics Processors},
    Institution = {EECS Department, University of California, Berkeley},
    Year = {2008},
    Month = {Feb},
    URL = {http://www.eecs.berkeley.edu/Pubs/TechRpts/2008/EECS-2008-11.html},
    Number = {UCB/EECS-2008-11},
    Abstract = {Recent developments in programmable, highly parallel Graphics Processing Units (GPUs) have enabled high performance implementations of machine learning algorithms.  We describe a solver for Support Vector Machine training, using Platt's Sequential Minimal Optimization algorithm, which achieves speedups of 5-32x over LibSVM running on a high-end traditional processor.  We also present a system for SVM classification which achieves speedups of 120-150x over LibSVM.}
}

EndNote citation:

%0 Report
%A Catanzaro, Bryan Christopher
%A Sundaram, Narayanan
%A Keutzer, Kurt
%T Fast Support Vector Machine Training and Classification on Graphics Processors
%I EECS Department, University of California, Berkeley
%D 2008
%8 February 8
%@ UCB/EECS-2008-11
%U http://www.eecs.berkeley.edu/Pubs/TechRpts/2008/EECS-2008-11.html
%F Catanzaro:EECS-2008-11