unbalanced: The package implements different data-driven method for unbalanced datasets

A dataset is said to be unbalanced when the class of interest (minority class) is much rarer than normal behaviour (majority class). The cost of missing a minority class is typically much higher that missing a majority class. Most learning systems are not prepared to cope with unbalanced data. Proposed strategies essentially belong to the following categories: sampling and distance-based. Sampling techniques up-sample or down-sample a class of instances. SMOTE generates synthetic minority examples. Distance based techniques use distances between input points to under-sample or to remove noisy and borderline examples.

Version: 1.1
Depends: FNN, RANN
Published: 2014-02-06
Author: Andrea Dal Pozzolo, Olivier Caelen and Gianluca Bontempi
Maintainer: Andrea Dal Pozzolo <adalpozz at ulb.ac.be>
License: GPL (≥ 3)
URL: http://mlg.ulb.ac.be
NeedsCompilation: no
CRAN checks: unbalanced results


Reference manual: unbalanced.pdf
Package source: unbalanced_1.1.tar.gz
Windows binaries: r-devel: unbalanced_1.1.zip, r-release: unbalanced_1.1.zip, r-oldrel: unbalanced_1.1.zip
OS X Snow Leopard binaries: r-release: unbalanced_1.1.tgz, r-oldrel: unbalanced_1.1.tgz
OS X Mavericks binaries: r-release: unbalanced_1.1.tgz
Old sources: unbalanced archive