smoothedLasso: A Framework to Smooth L1 Penalized Regression Operators using Nesterov Smoothing

We provide full functionality to smooth L1 penalized regression operators and to compute regression estimates thereof. For this, the objective function of a user-specified regression operator is first smoothed using Nesterov smoothing (see Y. Nesterov (2005) <doi:10.1007/s10107-004-0552-5>), resulting in a modified objective function with explicit gradients everywhere. The smoothed objective function and its gradient are minimized via BFGS, and the obtained minimizer is returned. Using Nesterov smoothing, the smoothed objective function can be made arbitrarily close to the original (unsmoothed) one. In particular, the Nesterov approach has the advantage that it comes with explicit accuracy bounds, both on the L1/L2 difference of the unsmoothed to the smoothed objective functions as well as on their respective minimizers (see G. Hahn, S.M. Lutz, N. Laha, C. Lange (2020) <doi:10.1101/2020.09.17.301788>). A progressive smoothing approach is provided which iteratively smoothes the objective function, resulting in more stable regression estimates.

Version: 1.5
Imports: Rdpack, Matrix
Published: 2020-10-08
Author: Georg Hahn [aut,cre], Sharon M. Lutz [ctb], Nilanjana Laha [ctb], Christoph Lange [ctb]
Maintainer: Georg Hahn <ghahn at hsph.harvard.edu>
License: GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
NeedsCompilation: no
CRAN checks: smoothedLasso results

Downloads:

Reference manual: smoothedLasso.pdf
Package source: smoothedLasso_1.5.tar.gz
Windows binaries: r-devel: smoothedLasso_1.5.zip, r-release: smoothedLasso_1.5.zip, r-oldrel: smoothedLasso_1.5.zip
macOS binaries: r-release: smoothedLasso_1.5.tgz, r-oldrel: smoothedLasso_1.5.tgz
Old sources: smoothedLasso archive

Linking:

Please use the canonical form https://CRAN.R-project.org/package=smoothedLasso to link to this page.