ldbod: Local Density-Based Outlier Detection

Flexible procedures to compute local density-based outlier scores for ranking outliers. Both exact and approximate nearest neighbor search can be implemented, while also accommodating multiple k values and four different local density-based methods. It allows for referencing a random subsample of input data or a user specified reference data set to compute outlier scores against, so both unsupervised and semi-supervised outlier detection can be implemented.

Version: 0.1.0
Depends: R (≥ 3.1.0)
Imports: stats, RANN, mnormt
Published: 2016-09-14
Author: Kristopher Williams
Maintainer: Kristopher Williams <kristopher.williams83 at gmail.com>
License: GPL-3
URL: https://github.com/kwilliams83/ldbod
NeedsCompilation: no
Materials: README
CRAN checks: ldbod results

Downloads:

Reference manual: ldbod.pdf
Package source: ldbod_0.1.0.tar.gz
Windows binaries: r-devel: ldbod_0.1.0.zip, r-release: ldbod_0.1.0.zip, r-oldrel: ldbod_0.1.0.zip
OS X Mavericks binaries: r-release: ldbod_0.1.0.tgz, r-oldrel: ldbod_0.1.0.tgz

Linking:

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