Presentation 2016-01-19
GPGPU Implementation of the MSD Method for Outlier Detection and Its Experimental Evaluation
Shotaro Asano, Masato Inagi, Shinobu Nagayama, Shin'ichi Wakabayashi,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In recent years,as the information,communication and sensing technologies advance,data streams have been continuously growing in size.To detect data which do not match expected patterns from such a large set of data,outlier detection methods have been being studied.As one of them,the MSD method,which extends a simple outlier detection method based on the Mahalanobis distance,has been proposed.It requires a larger computation time than the simple method,but it realizes more robust outlier detection.To ease the problem of large computation time,in this study,we propose a GPGPU implementation of the MSD method,and perform experimentalevaluations.In the MSD method,before calculating the Mahalanobis distance between the past and new data,the weights of some of the past data which are likely to be outliers are reduced in order to improve the robustness of the detection.In this study,we focus on a fact that it consists of independent processes of randomly generating multiple bases,and thus implement it so that these processes are executed in parallel.In experiments,we confirmed the effectiveness of our proposed implementation.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Outlier detection / Mahalanobis distance / MSD method / GPU
Paper # VLD2015-83,CPSY2015-115,RECONF2015-65
Date of Issue 2016-01-12 (VLD, CPSY, RECONF)

Conference Information
Committee VLD / CPSY / RECONF / IPSJ-SLDM / IPSJ-ARC
Conference Date 2016/1/19(3days)
Place (in Japanese) (See Japanese page)
Place (in English) Hiyoshi Campus, Keio University
Topics (in Japanese) (See Japanese page)
Topics (in English) FPGA Applications, etc
Chair Yusuke Matsunaga(Kyushu Univ.) / Yasuhiko Nakashima(NAIST) / Minoru Watanabe(Shizuoka Univ.) / Masahiro Fukui(Ritsumeikan Univ.) / Masahiro Goshima(国情研)
Vice Chair Takashi Takenana(NEC) / Koji Nakano(Hiroshima Univ.) / Hidetsugu Irie(Univ. of Tokyo) / Masato Motomura(Hokkaido Univ.) / Yuichiro Shibata(Nagasaki Univ.)
Secretary Takashi Takenana(Ritsumeikan Univ.) / Koji Nakano(Fujitsu Labs.) / Hidetsugu Irie(Fujitsu Labs.) / Masato Motomura(NII) / Yuichiro Shibata(Toshiba) / (Univ. of Tsukuba) / (Sharp)
Assistant Ittetsu Taniguchi(Ritsumeikan Univ.) / Shinya Takameda(NAIST) / Takeshi Ohkawa(Utsunomiya Univ.) / Kazuya Tanikagawa(Hiroshima City Univ.) / Takefumi Miyoshi(e-trees.Japan)

Paper Information
Registration To Technical Committee on VLSI Design Technologies / Technical Committee on Computer Systems / Technical Committee on Reconfigurable Systems / Special Interest Group on System and LSI Design Methodology / Special Interest Group on System Architecture
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) GPGPU Implementation of the MSD Method for Outlier Detection and Its Experimental Evaluation
Sub Title (in English)
Keyword(1) Outlier detection
Keyword(2) Mahalanobis distance
Keyword(3) MSD method
Keyword(4) GPU
1st Author's Name Shotaro Asano
1st Author's Affiliation Hiroshima City University(Hiroshima City Univ.)
2nd Author's Name Masato Inagi
2nd Author's Affiliation Hiroshima City University(Hiroshima City Univ.)
3rd Author's Name Shinobu Nagayama
3rd Author's Affiliation Hiroshima City University(Hiroshima City Univ.)
4th Author's Name Shin'ichi Wakabayashi
4th Author's Affiliation Hiroshima City University(Hiroshima City Univ.)
Date 2016-01-19
Paper # VLD2015-83,CPSY2015-115,RECONF2015-65
Volume (vol) vol.115
Number (no) VLD-398,CPSY-399,RECONF-400
Page pp.pp.37-42(VLD), pp.37-42(CPSY), pp.37-42(RECONF),
#Pages 6
Date of Issue 2016-01-12 (VLD, CPSY, RECONF)