RPEClust: Random Projection Ensemble Clustering Algorithm

Implements the methodology proposed by Anderlucci, Fortunato and Montanari (2019) <arXiv:1909.10832> for high-dimensional unsupervised classification. The random projection ensemble clustering algorithm applies a Gaussian Mixture Model to different random projections of the high-dimensional data and selects a subset of solutions accordingly to the Bayesian Information Criterion, computed here as discussed in Raftery and Dean (2006) <doi:10.1198/016214506000000113>. The clustering results obtained on the selected projections are then aggregated via consensus to derive the final partition.

Version: 0.1.0
Depends: R (≥ 3.6.0), clusteval
Imports: mclust, clue
Published: 2019-11-06
Author: L. Anderlucci [aut], F. Fortunato [aut, cre], A. Montanari [ctb]
Maintainer: Francesca Fortunato <francesca.fortunato3 at unibo.it>
License: GPL-3
URL: https://arxiv.org/abs/1909.10832
NeedsCompilation: no
CRAN checks: RPEClust results


Reference manual: RPEClust.pdf
Package source: RPEClust_0.1.0.tar.gz
Windows binaries: r-devel: RPEClust_0.1.0.zip, r-devel-gcc8: RPEClust_0.1.0.zip, r-release: RPEClust_0.1.0.zip, r-oldrel: not available
OS X binaries: r-release: RPEClust_0.1.0.tgz, r-oldrel: not available


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