Presentation 2005/3/23
Online Variable Selection From Multi-Dimentional Successively Data
Ryuji OSHIMA, Koichiro YAMAUCHI, Takashi OMORI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Variable selection from multi-dimensional inputs is very important issue for effective machine learning. Although, many any researchers have proposed various types of techniques to achieve the variable selection, almost all of them are based on off-line learning manner. In this paper, we propose a new paradigm to achieve quick variable selection. The new method generates various candidates selecting various variable subset which is likely to be useful. Then, the candidates are evaluated simultaneously, and the important variables are determined according to prediction accuracy. In the computer simulation, we show that the system realizes reasonable online variable selection.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) real time / variable selection / Online learning / Nearest Neighbor
Paper # NC2004-223
Date of Issue

Conference Information
Committee NC
Conference Date 2005/3/23(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) Online Variable Selection From Multi-Dimentional Successively Data
Sub Title (in English)
Keyword(1) real time
Keyword(2) variable selection
Keyword(3) Online learning
Keyword(4) Nearest Neighbor
1st Author's Name Ryuji OSHIMA
1st Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University()
2nd Author's Name Koichiro YAMAUCHI
2nd Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University
3rd Author's Name Takashi OMORI
3rd Author's Affiliation Graduate School of Information Science and Technology, Hokkaido University
Date 2005/3/23
Paper # NC2004-223
Volume (vol) vol.104
Number (no) 760
Page pp.pp.-
#Pages 5
Date of Issue