PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Discussion to least angle regression
Pascal Massart and Jean-Michel Loubes
Annals of Statistics Volume 32, Number 2, pp. 476-482, 2004.

Abstract

This paper is a discussion about the least angle regression algorithm proposed by Efron et al. In particular, we interprete their Mallows type model selection criterion as a least squares penalized criterion with a random penalty and show the connection with the works of Birgé and Massart on the topic.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
Theory & Algorithms
ID Code:1756
Deposited By:Pascal Massart
Deposited On:28 November 2005