smurf

Overview

smurf contains the implementation of the SMuRF (Sparse Multi-type Regularized Feature modeling; see Devriendt et al., 2021) algorithm to fit generalized linear models (GLMs) with multiple types of predictors via regularized maximum likelihood. Next to the fitting procedure, following functionality is available:

• Selection of the regularization tuning parameter lambda using three different approaches: in-sample, out-of-sample or using cross-validation.
• S3 methods to handle the fitted object including visualization of the coefficients and a model summary.

Documentation

More details on how to use the package functions can be found in the vignette or in the package documentation.

Installation

The smurf package can be installed from CRAN.

You can also install the latest development version of smurf from GitLab. If you work on Windows, make sure first that Rtools is installed. Then, install the latest development version of smurf using

install.packages("remotes")

remotes::install_git("https://gitlab.com/TReynkens/smurf.git")

Authors

Tom Reynkens, Sander Devriendt and Katrien Antonio

Reference

Devriendt, S., Antonio, K., Reynkens, T. and Verbelen, R. (2021). “Sparse Regression with Multi-type Regularized Feature Modeling”, Insurance: Mathematics and Economics, 96, 248–261. DOI:10.1016/j.insmatheco.2020.11.010.