Presentation 2004/9/21
Efficient Forecast Process for Nationwide Traffic Information Service
Masatoshi KUMAGAI, Takumi FUSHIKI, Takayoshi YOKOTA, Kazuya KIMITA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) This paper discusses about two extended methods of "Feature Space Forecast Method" for long-term traffic information forecasting. The first one is the method which decreases the size of a forecasting database one-tenth as large as that of usual regression analysis, by performing forecast process in the feature space shared by some links. The second one makes the best use of traffic information with a large number of missing data, by applying "Principal Component Analysis with Missing Data" for the process of feature space projection. It provides accurate forecast information even from probe-car data whose 80% of time series is missing.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Traffic Information / Forecasting / Principal Component Analysis / Feature Space / Missing Data
Paper # ITS2004-20
Date of Issue

Conference Information
Committee ITS
Conference Date 2004/9/21(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 Intelligent Transport Systems Technology (ITS)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Efficient Forecast Process for Nationwide Traffic Information Service
Sub Title (in English)
Keyword(1) Traffic Information
Keyword(2) Forecasting
Keyword(3) Principal Component Analysis
Keyword(4) Feature Space
Keyword(5) Missing Data
1st Author's Name Masatoshi KUMAGAI
1st Author's Affiliation Hitachi Research Lab., Hitachi Ltd.()
2nd Author's Name Takumi FUSHIKI
2nd Author's Affiliation Hitachi Research Lab., Hitachi Ltd.
3rd Author's Name Takayoshi YOKOTA
3rd Author's Affiliation Hitachi Research Lab., Hitachi Ltd.
4th Author's Name Kazuya KIMITA
4th Author's Affiliation Hitachi Research Lab., Hitachi Ltd.
Date 2004/9/21
Paper # ITS2004-20
Volume (vol) vol.104
Number (no) 325
Page pp.pp.-
#Pages 8
Date of Issue