NBLDA: Negative Binomial Linear Discriminant Analysis

We proposed a package for classification task which uses Negative Binomial distribution within Linear Discriminant Analysis (NBLDA). It is basically an extension of 'PoiClaClu' package to Negative Binomial distribution. The classification algorithms are based on the papers Dong et al. (2016, ISSN: 1471-2105) and Witten, DM (2011, ISSN: 1932-6157) for NBLDA and PLDA respectively. Although PLDA is a sparse algorithm and can be used for variable selection, the algorithm proposed by Dong et. al. is not sparse, hence, it uses all variables in the classifier. Here, we extent Dong et. al.'s algorithm to sparse case by shrinking overdispersion towards 0 (Yu et. al., 2013, ISSN: 1367-4803) and offset parameter towards 1 (as proposed by Witten DM, 2011). We support only the classification task with this version. However, the clustering task will be included with the following versions.

Version: 0.99.0
Depends: ggplot2
Imports: methods, stats, graphics, sSeq
Suggests: knitr
Published: 2018-07-01
Author: Dincer Goksuluk [aut, cre], Gokmen Zararsiz [aut], Selcuk Korkmaz [aut], Ahmet Ergun Karaagaoglu [ths, aut]
Maintainer: Dincer Goksuluk <dincer.goksuluk at hacettepe.edu.tr>
License: GPL-2 | GPL-3 [expanded from: GPL (≥ 2)]
NeedsCompilation: no
CRAN checks: NBLDA results

Downloads:

Reference manual: NBLDA.pdf
Package source: NBLDA_0.99.0.tar.gz
Windows binaries: r-devel: NBLDA_0.99.0.zip, r-release: NBLDA_0.99.0.zip, r-oldrel: NBLDA_0.99.0.zip
OS X binaries: r-release: not available, r-oldrel: not available

Linking:

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