Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Look-Ahead Screening Rules for the Lasso

Larsson, Johan LU orcid (2021) 22nd European Young Statisticians Meeting
p.61-65
Abstract
The lasso is a popular method to induce shrinkage and sparsity in the solution vector (coefficients) of regression problems, particularly when there are many predictors relative to the number of observations. Solving the lasso in this high-dimensional setting can, however, be computationally demanding. Fortunately, this demand can be alleviated via the use of screening rules that discard predictors prior to fitting the model, leading to a reduced problem to be solved. In this paper, we present a new screening strategy: look-ahead screening. Our method uses safe screening rules to find a range of penalty values for which a given predictor cannot enter the model, thereby screening predictors along the remainder of the path. In experiments we... (More)
The lasso is a popular method to induce shrinkage and sparsity in the solution vector (coefficients) of regression problems, particularly when there are many predictors relative to the number of observations. Solving the lasso in this high-dimensional setting can, however, be computationally demanding. Fortunately, this demand can be alleviated via the use of screening rules that discard predictors prior to fitting the model, leading to a reduced problem to be solved. In this paper, we present a new screening strategy: look-ahead screening. Our method uses safe screening rules to find a range of penalty values for which a given predictor cannot enter the model, thereby screening predictors along the remainder of the path. In experiments we show that these look-ahead screening rules outperform the active warm-start version of the Gap Safe rules. (Less)
Please use this url to cite or link to this publication:
author
organization
publishing date
type
Chapter in Book/Report/Conference proceeding
publication status
published
subject
keywords
lasso, screening rules, safe screening rules
host publication
22nd European young statisticians meeting - proceedings
editor
Makridis, Andreas ; Milienos, Fotios S. ; Papastamoulis, Panagiotis ; Parpoula, Christina and Rakitzis, Athanasios
pages
5 pages
publisher
Panteion University of Social and Political Sciences
conference name
22nd European Young Statisticians Meeting<br/>
conference location
Athens (Online)
conference dates
2021-09-06 - 2021-09-10
ISBN
978-960-7943-23-1
project
Optimization and Algorithms in Sparse Regression: Screening Rules, Coordinate Descent, and Normalization
language
English
LU publication?
yes
id
00d165b7-2f26-445d-8293-267ba8d3ac61
date added to LUP
2024-05-13 09:33:27
date last changed
2024-05-14 02:41:45
@inproceedings{00d165b7-2f26-445d-8293-267ba8d3ac61,
  abstract     = {{The lasso is a popular method to induce shrinkage and sparsity in the solution vector (coefficients) of regression problems, particularly when there are many predictors relative to the number of observations. Solving the lasso in this high-dimensional setting can, however, be computationally demanding. Fortunately, this demand can be alleviated via the use of screening rules that discard predictors prior to fitting the model, leading to a reduced problem to be solved. In this paper, we present a new screening strategy: look-ahead screening. Our method uses safe screening rules to find a range of penalty values for which a given predictor cannot enter the model, thereby screening predictors along the remainder of the path. In experiments we show that these look-ahead screening rules outperform the active warm-start version of the Gap Safe rules.}},
  author       = {{Larsson, Johan}},
  booktitle    = {{22nd European young statisticians meeting - proceedings}},
  editor       = {{Makridis, Andreas and Milienos, Fotios S. and Papastamoulis, Panagiotis and Parpoula, Christina and Rakitzis, Athanasios}},
  isbn         = {{978-960-7943-23-1}},
  keywords     = {{lasso; screening rules; safe screening rules}},
  language     = {{eng}},
  month        = {{09}},
  pages        = {{61--65}},
  publisher    = {{Panteion University of Social and Political Sciences}},
  title        = {{Look-Ahead Screening Rules for the Lasso}},
  url          = {{https://lup.lub.lu.se/search/files/183398675/larsson_2021_look-ahead_screening_rules_for_the_lasso.pdf}},
  year         = {{2021}},
}