Presentation 2001/12/14
Speed-up of kNN Search and kNN Classification by Filtering Based on Dimensionality Reduction
Kohei INOUE, Kiichi URAHAMA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Acceleration methods are presented for k nearest neighbor (NN) search and kNN classification on the basis of dimensionality reduction of feature vectors. Upper and lower bounds of distance between data are derived for general norm. They include distance in low dimension. Computational costs of kNN search and classification are reduced by filtering data on the basis of the inequality of distance. As an example of high dimensional data, color histograms of images are examined. Experiments show that the filtering with both upper and lower bounds is faster than that with only lower bound, and kNN classification which utilizes properties of classification is faster than that utilizes kNN search simply.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) kNN search / kNN classification / upper and lower bounds of distance / filtering
Paper # PRMU2001-176
Date of Issue

Conference Information
Committee PRMU
Conference Date 2001/12/14(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Pattern Recognition and Media Understanding (PRMU)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Speed-up of kNN Search and kNN Classification by Filtering Based on Dimensionality Reduction
Sub Title (in English)
Keyword(1) kNN search
Keyword(2) kNN classification
Keyword(3) upper and lower bounds of distance
Keyword(4) filtering
1st Author's Name Kohei INOUE
1st Author's Affiliation Faculty of Visual communication Design, Kyushu Institute of Design()
2nd Author's Name Kiichi URAHAMA
2nd Author's Affiliation Faculty of Visual communication Design, Kyushu Institute of Design
Date 2001/12/14
Paper # PRMU2001-176
Volume (vol) vol.101
Number (no) 525
Page pp.pp.-
#Pages 6
Date of Issue