Linear Regression in High Dimension and/or for Correlated Inputs. Issue 66 (23rd January 2015)

• Record Type: Journal Article
• Authors: Jacques, J.; Fraix-Burnet, D.
• Editors: Fraix-Burnet, D.; Valls-Gabaud, D.
• Is Part Of: EAS publications series. Issue 66(2014)
• Page Start: 149
• Page End: 165
• Publication Date: 2015-01-23
• Subjects: Astronomy -- Periodicals; 520
• DOI: 10.1051/eas/1466011
• Languages: English
• ISSNs: 1633-4760

Abstract

Abstract : Ordinary least square is the common way to estimate linear regression models. When inputs are correlated or when they are too numerous, regression methods using derived inputs directions or shrinkage methods can be efficient alternatives. Methods using derived inputs directions build new uncorrelated variables as linear combination of the initial inputs, whereas shrinkage methods introduce regularization and variable selection by penalizing the usual least square criterion. Both kinds of methods are presented and illustrated thanks to theR software on an astronomical dataset.