Selection of a threshold parameter based on the Quantile Universal Threshold (QUT) for GLM-lasso and Square-root lasso to obtain a sparse model with a good compromise between high true positive rate and low false discovery rate.
| Version: | 1.2 |
| Depends: | Matrix, glmnet, lars, flare |
| Published: | 2017-03-10 |
| Author: | Jairo Diaz Rodriguez, Sylvain Sardy, Caroline Giacobino, Nick Hengartner. |
| Maintainer: | Jairo Diaz Rodriguez <jairo.diaz at unige.ch> |
| License: | GPL-2 |
| NeedsCompilation: | no |
| CRAN checks: | qut results |
| Reference manual: | qut.pdf |
| Package source: | qut_1.2.tar.gz |
| Windows binaries: | r-devel: qut_1.2.zip, r-release: qut_1.2.zip, r-oldrel: qut_1.2.zip |
| OS X El Capitan binaries: | r-release: qut_1.2.tgz |
| OS X Mavericks binaries: | r-oldrel: qut_1.2.tgz |
| Old sources: | qut archive |
Please use the canonical form https://CRAN.R-project.org/package=qut to link to this page.