RT Journal Article
T1 A critical review of LASSO and its derivatives for variable selection under dependence among covariates
A1 Freijeiro-González, Laura
A1 Febrero Bande, Manuel
A1 González Manteiga, Wenceslao
K1 Covariates selection
K1 p > n
K1 L1 regularisation techniques
K1 LASSO
AB The limitations of the well-known LASSO regression as a variable selector are tested when there exists dependence structures among covariates. We analyse both the classic situation with n ≥ p and the high dimensional framework with p > n. Known restrictive properties of this methodology to guarantee optimality, as well as inconveniences in practice, are analysed and tested by means of an extensive simulation study. Examples of these drawbacks are showed making use of different dependence scenarios. In order to search for improvements, a broad comparison with LASSO derivatives and alternatives is carried out. Eventually, we give some guidance about what procedures work best in terms of the considered data nature
PB Wiley
YR 2022
FD 2022
LK http://hdl.handle.net/10347/29076
UL http://hdl.handle.net/10347/29076
LA eng
NO International Statistical Review (2022), 90, 1, 118–145. https://doi.org/10.1111/insr.12469
NO This work has been partially supported by the Spanish Ministerio de Economía, Industria y Competitividad grant MTM2016-76969-P, Xunta de Galicia Competitive Reference Groups 2017-2020 (ED431C 2017/38) and the Xunta de Galicia grant ED481A-2018/264
DS Minerva
RD 25 abr 2026