Accelerating exact similarity search on CPU-GPU systems

Takazumi Matsumoto, Man Lung Yiu

Research output: Chapter in book / Conference proceedingConference article published in proceeding or bookAcademic researchpeer-review

Abstract

In recent years, the use of Graphics Processing Units (GPUs) for data mining tasks has become popular. With modern processors integrating both CPUs and GPUs, it is also important to consider what tasks benefit from GPU processing and which do not, and apply a heterogeneous processing approach to improve the efficiency where applicable. Similarity search, also known as k-nearest neighbor search, is a key part of data mining applications and is used also extensively in applications such as multimedia search, where only a small subset of possible results are used. Our contribution is a new exact kNN algorithm with a compressed partial heapsort that outperforms other state-of-the-art exact kNN algorithms by leveraging both the GPU and CPU.
Original languageEnglish
Title of host publicationProceedings - 15th IEEE International Conference on Data Mining, ICDM 2015
PublisherIEEE
Pages320-329
Number of pages10
Volume2016-January
ISBN (Electronic)9781467395038
DOIs
Publication statusPublished - 5 Jan 2016
Event15th IEEE International Conference on Data Mining, ICDM 2015 - Atlantic City, United States
Duration: 14 Nov 201517 Nov 2015

Conference

Conference15th IEEE International Conference on Data Mining, ICDM 2015
Country/TerritoryUnited States
CityAtlantic City
Period14/11/1517/11/15

Keywords

  • Data mining
  • GPU
  • Heapsort
  • Heterogeneous processing
  • Nearest neighbor
  • Parallel
  • Similarity search

ASJC Scopus subject areas

  • Engineering(all)

Cite this