iFes: Incremental Feature Selection Algorithm accelerated by GPU

Incremental gene selection algorithm(iFes) expands the screening targets in each step of the forward stepwise algorithm, and proposes a novel feature selection algorithm iFes with both high accuracies and low feature numbers. Logistic regression classification algorithm accelerated by NVIDIA GPU is utilized to measure how good the currently chosen feature subset is.

Version: 1.0
Depends: R (≥ 2.10), ROCR, parallel, doParallel, foreach
Published: 2014-12-09
Author: Qinghan Meng, Fengfeng Zhou
Maintainer: Qinghan Meng <qinghan.meng at gmail.com>
License: GPL-2
NeedsCompilation: yes
SystemRequirements: Nvidia's CUDA toolkit (>= release 5.5)
Materials: INSTALL
CRAN checks: iFes results


Reference manual: iFes.pdf
Package source: iFes_1.0.tar.gz
Windows binaries: r-devel: not available, r-release: not available, r-oldrel: not available
OS X Snow Leopard binaries: r-release: not available, r-oldrel: not available
OS X Mavericks binaries: r-release: not available