Presentation 2012-11-08
An Efficient Input Variable Selection for a Linear Regression Model by NC Spectral Clustering
Koichi FUJIWARA, Hiroshi SAWADA, Manabu KANO,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Linear regression models have been widely accepted in many scientific and engineering fields for the estimation or interpretation of phenomena. When a linear regression model is built, appropriate input variables have to be selected to achieve high estimation performance. This work proposes new methodologies for selecting input variables for linear regression models using nearest correlation spectral clustering (NCSC), which is a correlation-based clustering method. In the present work, NCSC is used for variable group construction, and a few variable groups are selected by their contribution to estimates; it is referred to as NCSC-based variable selection (NCSC-VS). The usefulness of the proposed NCSC-VS is demonstrated through an industrial application to a chemical process.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Linear regression / Variable selection / Spectral clustering / Partial least squares
Paper # IBISML2012-84
Date of Issue

Conference Information
Committee IBISML
Conference Date 2012/10/31(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 Information-Based Induction Sciences and Machine Learning (IBISML)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) An Efficient Input Variable Selection for a Linear Regression Model by NC Spectral Clustering
Sub Title (in English)
Keyword(1) Linear regression
Keyword(2) Variable selection
Keyword(3) Spectral clustering
Keyword(4) Partial least squares
1st Author's Name Koichi FUJIWARA
1st Author's Affiliation Kyoto University()
2nd Author's Name Hiroshi SAWADA
2nd Author's Affiliation NTT NTT Communication Science Laboratories
3rd Author's Name Manabu KANO
3rd Author's Affiliation Kyoto University
Date 2012-11-08
Paper # IBISML2012-84
Volume (vol) vol.112
Number (no) 279
Page pp.pp.-
#Pages 8
Date of Issue