fmf: Fast Class Noise Detector with Multi-Factor-Based Learning (original) (raw)

A fast class noise detector which provides noise score for each observations. The package takes advantage of 'RcppArmadillo' to speed up the calculation of distances between observations.

Version:

1.1.1

Depends:

R (≥ 2.10.0)

Imports:

Rcpp, caret, solitude, kernlab, C50, e1071, FactoMineR, dplyr, factoextra, ggplot2

LinkingTo:

Rcpp, RcppArmadillo

Suggests:

testthat, covr, knitr, rmarkdown

Published:

2020-09-03

DOI:

10.32614/CRAN.package.fmf

Author:

Wanwan Zheng [aut, cre], Mingzhe Jin [aut], Lampros Mouselimis [ctb, cph]

Maintainer:

Wanwan Zheng

License:

MIT + file

NeedsCompilation:

yes

SystemRequirements:

libarmadillo: apt-get install -y libarmadillo-dev (deb), libblas: apt-get install -y libblas-dev (deb), liblapack: apt-get install -y liblapack-dev (deb), libarpack++2: apt-get install -y libarpack++2-dev (deb), gfortran: apt-get install -y gfortran (deb)

CRAN checks:

fmf results [issues need fixing before 2025-11-15]