Skip to content

Datasets

OutlierDetectionData provides access to collections of outlier detection datasets. The following collections are currently supported:

  • ODDS, Outlier Detection DataSets, Shebuti Rayana, 2016
  • ELKI, On the Evaluation of Unsupervised Outlier Detection, Campos et al., 2016
  • TSAD, The UCR Time Series Archive, Dau et al., 2018

For the TSAD collection, the class with the least members is chosen as the anomaly class and all other classes are defined as normal. If there are multiple classes, the lexically first class is chosen.

The following methods are defined for all collections.

list

# OutlierDetectionData.list — Function.

list([prefix])

List the available datasets in a dataset collection optionally given a prefix.

Parameters

prefix::Union{Regex, AbstractString}

Regex or string used to filter the datasets.

load

# OutlierDetectionData.load — Function.

load(dataset)

Load a given dataset from a dataset collection.

Parameters

name::AbstractString

Name of the dataset to load.