Linearized Bregman Algorithms for Generalized Linear Models

Efficient procedures for fitting the regularization path for linear, binomial, multinomial, Ising and Potts models with lasso, group lasso or column lasso(only for multinomial) penalty. The package uses Linearized Bregman Algorithm to solve the regularization path through iterations. Bregman Inverse Scale Space Differential Inclusion solver is also provided for linear model with lasso penalty.


News

Reference manual

It appears you don't have a PDF plugin for this browser. You can click here to download the reference manual.

install.packages("Libra")

1.6 by Jiechao Xiong, a year ago


http://arxiv.org/abs/1406.7728


Browse source code at https://github.com/cran/Libra


Authors: Feng Ruan , Jiechao Xiong and Yuan Yao


Documentation:   PDF Manual  


GPL-2 license


Depends on nnls

Suggests lars, MASS, igraph

System requirements: GNU Scientific Library (GSL)


See at CRAN