bsnsing: Build Decision Trees with Optimal Multivariate Splits

Functions for training an optimal decision tree classifier, making predictions and generating latex code for plotting. Works for two-class and multi-class classification problems. The algorithm seeks the optimal Boolean rule consisting of multiple variables to split a node, resulting in shorter trees. Use bsnsing() to build a tree, predict() to make predictions and plot() to plot the tree into latex and PDF. See Yanchao Liu (2022) <doi:10.48550/arXiv.2205.15263> for technical details. Source code and more data sets are at <https://github.com/profyliu/bsnsing/>.

Version: 1.0.1
Imports: methods, Rcpp (≥ 1.0.3)
LinkingTo: Rcpp
Suggests: C50, party, rpart, tree
Published: 2022-07-03
DOI: 10.32614/CRAN.package.bsnsing
Author: Yanchao Liu
Maintainer: Yanchao Liu <yanchaoliu at wayne.edu>
License: GPL-3
NeedsCompilation: yes
Materials: README
CRAN checks: bsnsing results

Documentation:

Reference manual: bsnsing.pdf

Downloads:

Package source: bsnsing_1.0.1.tar.gz
Windows binaries: r-devel: bsnsing_1.0.1.zip, r-release: bsnsing_1.0.1.zip, r-oldrel: bsnsing_1.0.1.zip
macOS binaries: r-release (arm64): bsnsing_1.0.1.tgz, r-oldrel (arm64): bsnsing_1.0.1.tgz, r-release (x86_64): bsnsing_1.0.1.tgz, r-oldrel (x86_64): bsnsing_1.0.1.tgz

Linking:

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