SLASSO: a scaled LASSO for multicollinear situations

Arashi M., ASAR Y., YÜZBAŞI B.

JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION, vol.91, no.15, pp.3170-3183, 2021 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 91 Issue: 15
  • Publication Date: 2021
  • Doi Number: 10.1080/00949655.2021.1924174
  • Journal Name: JOURNAL OF STATISTICAL COMPUTATION AND SIMULATION
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, Aerospace Database, Business Source Elite, Business Source Premier, CAB Abstracts, Communication Abstracts, Metadex, Veterinary Science Database, zbMATH, Civil Engineering Abstracts
  • Page Numbers: pp.3170-3183
  • Keywords: Biasing parameter, L-1-penalty, LASSO, Liu estimation, Multicollinearity, variable selection, BIASED-ESTIMATION, REGRESSION, REGULARIZATION, ESTIMATORS, SELECTION, RIDGE
  • Inonu University Affiliated: Yes

Abstract

We propose a re-scaled LASSO by pre-multiplying the LASSO with a matrix term, namely, scaled LASSO (SLASSO), for multicollinear situations. Our numerical study has shown that the SLASSO is comparable with other sparse modeling techniques and often outperforms the LASSO and elastic net. Our findings open new visions about using the LASSO still for sparse modeling and variable selection. We conclude our study by pointing that the same efficient algorithm can solve the SLASSO for solving the LASSO and suggest following the same construction technique for other penalized estimators