On the “degrees of freedom” of the lasso
Top Cited Papers
Open Access
- 1 October 2007
- journal article
- Published by Institute of Mathematical Statistics in The Annals of Statistics
- Vol. 35 (5) , 2173-2192
- https://doi.org/10.1214/009053607000000127
Abstract
We study the effective degrees of freedom of the lasso in the framework of Stein’s unbiased risk estimation (SURE). We show that the number of nonzero coefficients is an unbiased estimate for the degrees of freedom of the lasso—a conclusion that requires no special assumption on the predictors. In addition, the unbiased estimator is shown to be asymptotically consistent. With these results on hand, various model selection criteria—Cp, AIC and BIC—are available, which, along with the LARS algorithm, provide a principled and efficient approach to obtaining the optimal lasso fit with the computational effort of a single ordinary least-squares fit.Keywords
All Related Versions
This publication has 19 references indexed in Scilit:
- Statistical challenges with high dimensionality: feature selection in knowledge discoveryPublished by European Mathematical Society - EMS - Publishing House GmbH ,2007
- Optimal Model Assessment, Selection, and CombinationJournal of the American Statistical Association, 2006
- Can the strengths of AIC and BIC be shared? A conflict between model indentification and regression estimationBiometrika, 2005
- Adaptive Model SelectionJournal of the American Statistical Association, 2002
- A new approach to variable selection in least squares problemsIMA Journal of Numerical Analysis, 2000
- On Measuring and Correcting the Effects of Data Mining and Model SelectionJournal of the American Statistical Association, 1998
- Adapting to Unknown Smoothness via Wavelet ShrinkageJournal of the American Statistical Association, 1995
- Estimation of the Mean of a Multivariate Normal DistributionThe Annals of Statistics, 1981
- Estimating the Dimension of a ModelThe Annals of Statistics, 1978
- Some Comments onCpTechnometrics, 1973