solitude (version 0.1.3)

isolationForest: Grow an isolation forest

Description

Implements isolation forest from Isolation-Based Anomaly Detection (<doi:10.1145/2133360.2133363>)

Usage

isolationForest(dataset, ...)

Arguments

dataset

Dataframe with column names

...

Optional arguments passed to ranger

Value

An object(list) of class 'solitude'. The isolationforest is in the 'forest' slot

Details

This function internally uses ranger. Arguments may be passed to it except these: dependent.variable.name, data, mtry, min.node.size, splitrule, num.random.splits, respect.unordered.factors. Pass 'seed' argument for reproducibility

Examples

# NOT RUN {
mo <- isolationForest(iris, seed = 1)
# }