picasso: Pathwise Calibrated Sparse Shooting Algorithm

Implement a new family of efficient algorithms, called PathwIse CalibrAted Sparse Shooting AlgOrithm, for a variety of sparse learning problems, including Sparse Linear Regression, Sparse Logistic Regression, Sparse Column Inverse Operator and Sparse Multivariate Regression. Different types of active set identification schemes are implemented, such as cyclic search, greedy search, stochastic search and proximal gradient search. Besides, the package provides the choices between convex (L1 norm) and non-convex (MCP and SCAD) regularizations. Moreover, group regularization, such as group Lasso, group MCP and group SCAD, are also implemented for Sparse Linear Regression, Sparse Logistic Regression and Sparse Multivariate Regression.

Version: 0.2.0
Depends: R (≥ 2.15.0), lattice, igraph, MASS, Matrix
Published: 2014-10-06
Author: Xingguo Li, Tuo Zhao and Han Liu
Maintainer: Xingguo Li <xingguo.leo at gmail.com>
License: GPL-2
NeedsCompilation: yes
CRAN checks: picasso results

Downloads:

Reference manual: picasso.pdf
Package source: picasso_0.2.0.tar.gz
Windows binaries: r-devel: picasso_0.2.0.zip, r-release: picasso_0.2.0.zip, r-oldrel: picasso_0.2.0.zip
OS X Snow Leopard binaries: r-release: picasso_0.2.0.tgz, r-oldrel: picasso_0.2.0.tgz
OS X Mavericks binaries: r-release: picasso_0.2.0.tgz
Old sources: picasso archive