Spline Regression with Automatic Knot Selection

Abstract : In this paper we introduce a new method for automatically selecting knots in spline regression. The approach consists in setting a large number of initial knots and fitting the spline regression through a penalized likelihood procedure called adaptive ridge. The proposed method is similar to penalized spline regression methods (e.g. P-splines), with the noticeable difference that the output is a sparse spline regression with a small number of knots. We show that our method called A-spline, for adaptive splines yields sparse regression models with high interpretability, while having similar predictive performance similar to penalized spline regression methods. A-spline is applied both to simulated and real dataset. A fast and publicly available implementation in R is provided along with this paper.
Type de document :
Pré-publication, Document de travail
Liste complète des métadonnées

Contributeur : Vivien Goepp <>
Soumis le : vendredi 3 août 2018 - 13:32:15
Dernière modification le : vendredi 4 janvier 2019 - 17:33:38
Document(s) archivé(s) le : dimanche 4 novembre 2018 - 13:24:46


Fichiers produits par l'(les) auteur(s)


  • HAL Id : hal-01853459, version 1
  • ARXIV : 1808.01770


Vivien Goepp, Olivier Bouaziz, Grégory Nuel. Spline Regression with Automatic Knot Selection. 2018. 〈hal-01853459〉



Consultations de la notice


Téléchargements de fichiers