ExhaustiveSearch: A Fast and Scalable Exhaustive Feature Selection Framework

The goal of this package is to provide an easy to use, fast and scalable exhaustive search framework. Exhaustive feature selections typically require a very large number of models to be fitted and evaluated. Execution speed and memory management are crucial factors here. This package provides solutions for both. Execution speed is optimized by using a multi-threaded C++ backend, and memory issues are solved by by only storing the best results during execution and thus keeping memory usage constant.

Version: 1.0.1
Imports: Rcpp
LinkingTo: Rcpp, RcppArmadillo
Suggests: mlbench
Published: 2021-01-18
Author: Rudolf Jagdhuber [aut, cre], Jorge Nocedal [cph] (lbfgs c library), Naoaki Okazaki [cph] (lbfgs c library)
Maintainer: Rudolf Jagdhuber <r.jagdhuber at gmail.com>
BugReports: https://github.com/RudolfJagdhuber/ExhaustiveSearch/issues
License: GPL (≥ 3)
URL: https://github.com/RudolfJagdhuber/ExhaustiveSearch
NeedsCompilation: yes
Materials: README NEWS
CRAN checks: ExhaustiveSearch results

Documentation:

Reference manual: ExhaustiveSearch.pdf

Downloads:

Package source: ExhaustiveSearch_1.0.1.tar.gz
Windows binaries: r-devel: ExhaustiveSearch_1.0.1.zip, r-release: ExhaustiveSearch_1.0.1.zip, r-oldrel: ExhaustiveSearch_1.0.1.zip
macOS binaries: r-release (arm64): ExhaustiveSearch_1.0.1.tgz, r-oldrel (arm64): ExhaustiveSearch_1.0.1.tgz, r-release (x86_64): ExhaustiveSearch_1.0.1.tgz
Old sources: ExhaustiveSearch archive

Linking:

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