xrf: eXtreme RuleFit

An implementation of the RuleFit algorithm as described in Friedman & Popescu (2008) <doi:10.1214/07-AOAS148>. eXtreme Gradient Boosting ('XGBoost') is used to build rules, and 'glmnet' is used to fit a sparse linear model on the raw and rule features. The result is a model that learns similarly to a tree ensemble, while often offering improved interpretability and achieving improved scoring runtime in live applications. Several algorithms for reducing rule complexity are provided, most notably hyperrectangle de-overlapping. All algorithms scale to several million rows and support sparse representations to handle tens of thousands of dimensions.

Version: 0.2.2
Depends: R (≥ 3.1.0)
Imports: Matrix, glmnet (≥ 3.0), xgboost (≥ 0.71.2), dplyr, fuzzyjoin, rlang, methods
Suggests: testthat, covr
Published: 2022-10-04
Author: Karl Holub [aut, cre]
Maintainer: Karl Holub <karljholub at gmail.com>
BugReports: https://github.com/holub008/xrf/issues
License: MIT + file LICENSE
URL: https://github.com/holub008/xrf
NeedsCompilation: no
Materials: README
CRAN checks: xrf results

Documentation:

Reference manual: xrf.pdf

Downloads:

Package source: xrf_0.2.2.tar.gz
Windows binaries: r-devel: xrf_0.2.2.zip, r-release: xrf_0.2.2.zip, r-oldrel: xrf_0.2.2.zip
macOS binaries: r-release (arm64): xrf_0.2.2.tgz, r-oldrel (arm64): xrf_0.2.2.tgz, r-release (x86_64): xrf_0.2.2.tgz
Old sources: xrf archive

Reverse dependencies:

Reverse suggests: butcher, rules

Linking:

Please use the canonical form https://CRAN.R-project.org/package=xrf to link to this page.