Linear Regression in High Dimension and/or for Correlated Inputs - Archive ouverte HAL
Chapitre D'ouvrage Année : 2014

Linear Regression in High Dimension and/or for Correlated Inputs

Résumé

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 the R software on an astronomical dataset.

Dates et versions

hal-01115109 , version 1 (10-02-2015)

Identifiants

Citer

Julien Jacques, Didier Fraix-Burnet. Linear Regression in High Dimension and/or for Correlated Inputs. D; Fraix-Burnet; D. Valls-Gabaud. Statistics for Astrophysics Methods and Applications of the Regression, 66, EDP Sciences, pp.149-165, 2014, EAS Publications Series, 978-2-7598-1729-0. ⟨10.1051/eas/1466011⟩. ⟨hal-01115109⟩

Relations

209 Consultations
0 Téléchargements

Altmetric

Partager

More