Regression Shrinkage and Selection Via the Lasso
Robert John Tibshirani · Journal of the Royal Statistical Society Series B (Statistical Methodology) · 1996
SUMMARY We propose a new method for estimation in linear models. The ‘lasso’ minimizes the residual sum of squares subject to the sum of the absolute value of the coefficients being less than a constant. Because of the nature of this constraint it tends to produce some coefficients that are exactly 0 and hence gives interpretable models. Our simulation studies suggest that the lasso enjoys some of the favourable properties of both subset selection and ridge regression. It produces interpretable models like subset selection and exhibits the stability of ridge regression. There is also an interesting relationship with recent work in adaptive function estimation by Donoho and Johnstone. The lasso idea is quite general and can be applied in a variety of statistical models: extensions to generalized regression models and tree-based models are briefly described.
Cited by 51,245 works. doi:10.1111/j.2517-6161.1996.tb02080.x
Authors
References
- An Introduction to the Bootstrap
- Classification and Regression Trees.
- Classification and Regression Trees.
- Bootstrap Methods: Another Look at the Jackknife
- An Introduction to the Bootstrap
- Generalized Additive Models.
- Solving Least Squares Problems
- Ideal Spatial Adaptation by Wavelet Shrinkage
- Multivariate Adaptive Regression Splines
- Practical optimization