evtree: Evolutionary Learning of Globally Optimal Trees

Commonly used classification and regression tree methods like the CART algorithm are recursive partitioning methods that build the model in a forward stepwise search. Although this approach is known to be an efficient heuristic, the results of recursive tree methods are only locally optimal, as splits are chosen to maximize homogeneity at the next step only. An alternative way to search over the parameter space of trees is to use global optimization methods like evolutionary algorithms. The 'evtree' package implements an evolutionary algorithm for learning globally optimal classification and regression trees in R. CPU and memory-intensive tasks are fully computed in C++ while the 'partykit' package is leveraged to represent the resulting trees in R, providing unified infrastructure for summaries, visualizations, and predictions.

Version: 1.0-8
Depends: R (≥ 3.3.0), partykit
Suggests: Formula, kernlab, lattice, mlbench, multcomp, party, rpart, xtable
Published: 2019-05-26
DOI: 10.32614/CRAN.package.evtree
Author: Thomas Grubinger [aut, cre], Achim Zeileis ORCID iD [aut], Karl-Peter Pfeiffer [aut]
Maintainer: Thomas Grubinger <ThomasGrubinger at gmail.com>
License: GPL-2 | GPL-3
NeedsCompilation: yes
Citation: evtree citation info
Materials: NEWS
In views: MachineLearning
CRAN checks: evtree results


Reference manual: evtree.pdf
Vignettes: Evolutionary Learning of Globally Optimal Classification and Regression Trees in R


Package source: evtree_1.0-8.tar.gz
Windows binaries: r-devel: evtree_1.0-8.zip, r-release: evtree_1.0-8.zip, r-oldrel: evtree_1.0-8.zip
macOS binaries: r-release (arm64): evtree_1.0-8.tgz, r-oldrel (arm64): evtree_1.0-8.tgz, r-release (x86_64): evtree_1.0-8.tgz, r-oldrel (x86_64): evtree_1.0-8.tgz
Old sources: evtree archive

Reverse dependencies:

Reverse imports: insurancerating, pheble
Reverse suggests: flowml, fscaret, mlr, r2pmml, stablelearner


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