A GPU Based SVM Method with Accelerated Kernel Matrix Calculation

Bo Yan, Yitian Ren, Zijiang Yang

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Citations (Scopus)

Abstract

Support vector machine (SVM) is a popular classifier dealing with small-scale datasets. It has outstanding performance compared to other classifiers. However the execution time is extremely long when training Big Data. The Graphics Processing Unit (GPU) is a massively parallel device which performs very well as a co-processor. NVIDIA proposed a programming platform, CUDA, in 2006, which makes it much easier to program on GPU for large-scale calculation. This paper proposes a GPU based accelerated SVM method. Matrix multiplication, which is the main procedure of kernel matrix calculation in the training stage, is the major issue for high time complexity. Thus, we propose a method to calculate the kernel matrix using GPU. The polynomial kernel function is applied in this paper. Experiment results indicate that the proposed method can increase the computation speed by as much as 41 times for the studied datasets.

Original languageEnglish
Title of host publicationProceedings - 2015 IEEE International Congress on Big Data, BigData Congress 2015
EditorsLatifur Khan, Carminati Barbara
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages41-46
Number of pages6
ISBN (Electronic)9781467372787
DOIs
Publication statusPublished - 17 Aug 2015
Event4th IEEE International Congress on Big Data, BigData Congress 2015 - New York City, United States
Duration: 27 Jun 20152 Jul 2015

Publication series

NameProceedings - 2015 IEEE International Congress on Big Data, BigData Congress 2015

Conference

Conference4th IEEE International Congress on Big Data, BigData Congress 2015
Country/TerritoryUnited States
CityNew York City
Period27/06/152/07/15

Keywords

  • Big Data
  • CUDA
  • cross validation
  • graphics processing unit
  • kernel matrix
  • support vector machine

Fingerprint

Dive into the research topics of 'A GPU Based SVM Method with Accelerated Kernel Matrix Calculation'. Together they form a unique fingerprint.

Cite this

Yan, B., Ren, Y., & Yang, Z. (2015). A GPU Based SVM Method with Accelerated Kernel Matrix Calculation. In L. Khan, & C. Barbara (Eds.), Proceedings - 2015 IEEE International Congress on Big Data, BigData Congress 2015 (pp. 41-46). Article 7207200 (Proceedings - 2015 IEEE International Congress on Big Data, BigData Congress 2015). Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/BigDataCongress.2015.16