PCR and PLS for Clusterwise Regression on Functional Data

Cristian Preda (1) , Gilbert Saporta (2)

1 MODAL - MOdel for Data Analysis and Learning
2 CEDRIC - MSDMA - CEDRIC. Méthodes statistiques de data-mining et apprentissage

Clusterwise regression is applied to functional data, using PCR and PLS as regularization methods for the functional linear regression model.We compare these two approaches on simulated data as well as on stock exchange data.

Mots clés: Clusterwise regression; Functional data

Mots clés: Données fonctionnelles; Régression clusterwise

Domaines: Statistiques [math.ST]

Cristian Preda, Gilbert Saporta. PCR and PLS for Clusterwise Regression on Functional Data. Brito P., Bertrand P., Cucumel G., De Carvalho F. (Eds.). Selected Contributions in Data Analysis and Classification, Springer, pp.589-598, 2007, 978-3-540-73558-8. ⟨hal-01125361⟩