In the framework of Generalized Additive Models (GAM) an automatic data-driven procedure is introduced for assigning an appropriate smoother to each covariate and for defining an ordering entrance for the covariates in the model. The resulting Smoothing Score algorithm aims to improve model indentifiability. It uses the bagging procedure in order to select the smoothers to be assigned to each covariate and a new scoring measure able to rank the candidate smoothers with respect to their bagged predictive accuracy. The adequacy of this scoring measure is evaluated on artificial data. A comparison between the smoothing score algorithm and the standard GAM is made using real data concerning a classification task.

Smoothing Score Algorithm for Generalized Additive Models

CONVERSANO, CLAUDIO
2004

Abstract

In the framework of Generalized Additive Models (GAM) an automatic data-driven procedure is introduced for assigning an appropriate smoother to each covariate and for defining an ordering entrance for the covariates in the model. The resulting Smoothing Score algorithm aims to improve model indentifiability. It uses the bagging procedure in order to select the smoothers to be assigned to each covariate and a new scoring measure able to rank the candidate smoothers with respect to their bagged predictive accuracy. The adequacy of this scoring measure is evaluated on artificial data. A comparison between the smoothing score algorithm and the standard GAM is made using real data concerning a classification task.
978-3-540-20889-1
GAM; Backfitting; Anova
File in questo prodotto:
File Dimensione Formato  
Conversano_Chiodi2004.pdf

Solo gestori archivio

Descrizione: DEF
Tipologia: versione editoriale
Dimensione 3.21 MB
Formato Adobe PDF
3.21 MB Adobe PDF   Visualizza/Apri   Richiedi una copia

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11584/20055
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
  • ???jsp.display-item.citation.isi??? 0
social impact