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 neighborhood sizes and four different local density-based methods. It allows for referencing a random subsample of the 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.
|Date of publication||2016-12-22 09:04:34|
|Maintainer||Kristopher Williams <firstname.lastname@example.org>|