xtune: Regularized Regression with Differential Penalties Integrating External Information

Extends standard penalized regression (Lasso and Ridge) to allow differential shrinkage based on external information with the goal of achieving a better prediction accuracy. Examples of external information include the grouping of predictors, prior knowledge of biological importance, external p-values, function annotations, etc. The choice of multiple tuning parameters is done using an Empirical Bayes approach. A majorization-minimization algorithm is employed for implementation.

Version: 0.1.0
Depends: R (≥ 2.10)
Imports: glmnet, stats, selectiveInference
Suggests: knitr, numDeriv, lbfgs, rmarkdown, testthat, covr
Published: 2019-05-24
Author: Chubing Zeng
Maintainer: Chubing Zeng <chubingz at usc.edu>
License: MIT + file LICENSE
NeedsCompilation: no
Materials: README
CRAN checks: xtune results


Reference manual: xtune.pdf
Vignettes: Vignette Title
Package source: xtune_0.1.0.tar.gz
Windows binaries: r-devel: xtune_0.1.0.zip, r-release: xtune_0.1.0.zip, r-oldrel: xtune_0.1.0.zip
macOS binaries: r-release: xtune_0.1.0.tgz, r-oldrel: xtune_0.1.0.tgz


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