Presentation 2003/5/22
Performance Analysis of a Parallel Law-of-the-Jungle Algorithm for Generating Codebooks of Vector Quantization
Shintaro MOMOSE, Kentaro SANO, Hiroyuki TAKIZAWA, Taira NAKAJIMA, Hiroaki KOBAYASHI, Tadao NAKAMURA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Vector quantization is an attractive technique for lossy data compression, which has been a key technology for efficient data storage andlor transfer. So far, various algorithms have been proposed to design optimal codebooks presenting quantization with minimized errors. In particular, the Law-of-the-Jungle(LOJ) learning algorithm has been proposed to achieve rapid codebook design by algorithmic improvements. However, its acceleration is still required when large data sets are processed on a single computer. In order to achieve faster codebook design, we have been proposed a scalable parallel codebook design algorithm for parallel computers. This paper analyzes and evaluates the performance of the parallel LOJ learning algorithm on three types of parallel computers: an IBM SP2, an NEC AzusA and a PC cluster.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Vector Quantization / Codebook generation / Parallel processing / LOJ algorithm
Paper # NC2003-5
Date of Issue

Conference Information
Committee NC
Conference Date 2003/5/22(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 Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Performance Analysis of a Parallel Law-of-the-Jungle Algorithm for Generating Codebooks of Vector Quantization
Sub Title (in English)
Keyword(1) Vector Quantization
Keyword(2) Codebook generation
Keyword(3) Parallel processing
Keyword(4) LOJ algorithm
1st Author's Name Shintaro MOMOSE
1st Author's Affiliation Graduate School of Information Sciences, Tohoku University()
2nd Author's Name Kentaro SANO
2nd Author's Affiliation Graduate School of Information Sciences, Tohoku University
3rd Author's Name Hiroyuki TAKIZAWA
3rd Author's Affiliation Information Synergy Center, Tohoku University
4th Author's Name Taira NAKAJIMA
4th Author's Affiliation Graduate School of Engineering
5th Author's Name Hiroaki KOBAYASHI
5th Author's Affiliation Information Synergy Center, Tohoku University
6th Author's Name Tadao NAKAMURA
6th Author's Affiliation Graduate School of Information Sciences, Tohoku University
Date 2003/5/22
Paper # NC2003-5
Volume (vol) vol.103
Number (no) 92
Page pp.pp.-
#Pages 6
Date of Issue