pre: Prediction Rule Ensembles

Derives prediction rule ensembles (PREs). Largely follows the procedure for deriving PREs as described in Friedman & Popescu (2008; <doi:10.1214/07-AOAS148>), with several adjustments and improvements. The main function pre() derives prediction rule ensembles consisting of rules and/or linear terms for binary, count and continuous responses. Function gpe() derives generalized prediction ensembles, consisting of rules, hinge and linear functions of the predictor variables.

Version: 0.4
Depends: R (≥ 3.1.0)
Imports: glmnet, MatrixModels, Formula, methods, graphics, partykit, earth, stringr
Suggests: akima, grid, foreach, testthat, mlbench, datasets
Published: 2017-08-31
Author: Marjolein Fokkema [aut, cre], Benjamin Christoffersen [aut]
Maintainer: Marjolein Fokkema <m.fokkema at>
License: GPL-2 | GPL-3
NeedsCompilation: no
Materials: README NEWS
CRAN checks: pre results


Reference manual: pre.pdf
Package source: pre_0.4.tar.gz
Windows binaries: r-devel:, r-release:, r-oldrel:
OS X El Capitan binaries: r-release: pre_0.4.tgz
OS X Mavericks binaries: r-oldrel: pre_0.4.tgz
Old sources: pre archive


Please use the canonical form to link to this page.