×

zbMATH — the first resource for mathematics

Unified LASSO estimation by least squares approximation. (English) Zbl 1306.62167
Summary: We propose a method of least squares approximation (LSA) for unified yet simple LASSO estimation. Our general theoretical framework includes ordinary least squares, generalized linear models, quantile regression, and many others as special cases. Specifically, LSA can transfer many different types of LASSO objective functions into their asymptotically equivalent least squares problems. Thereafter, the standard asymptotic theory can be established and the LARS algorithm can be applied. In particular, if the adaptive LASSO penalty and a Bayes information criterion-type tuning parameter selector are used, the resulting LSA estimator can be as efficient as the oracle. Extensive numerical studies confirm our theory.

MSC:
62J07 Ridge regression; shrinkage estimators (Lasso)
65C60 Computational problems in statistics (MSC2010)
Software:
R; S-PLUS
PDF BibTeX XML Cite
Full Text: DOI