grpnet: Group Elastic Net Regularized GLMs and GAMs

Efficient algorithms for fitting generalized linear and additive models with group elastic net penalties as described in Helwig (2024) <doi:10.1080/10618600.2024.2362232>. Implements group LASSO, group MCP, and group SCAD with an optional group ridge penalty. Computes the regularization path for linear regression (gaussian), logistic regression (binomial), multinomial logistic regression (multinomial), log-linear count regression (poisson and negative.binomial), and log-linear continuous regression (gamma and inverse gaussian). Supports default and formula methods for model specification, k-fold cross-validation for tuning the regularization parameters, and nonparametric regression via tensor product reproducing kernel (smoothing spline) basis function expansion.

Getting started

Package details

AuthorNathaniel E. Helwig [aut, cre]
MaintainerNathaniel E. Helwig <helwig@umn.edu>
LicenseGPL (>= 2)
Version0.4
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:
install.packages("grpnet")

Try the grpnet package in your browser

Any scripts or data that you put into this service are public.

grpnet documentation built on June 22, 2024, 9:35 a.m.