Four different outlier detection techniques are described in the following sections. They are as follows:
Detect Outlier (Distances)
Detect Outlier (Densities)
Detect Outlier (LOF)
Detect Outlier (COF)
None of these algorithms will automatically find the correct outliers for the data being explored. Given their parameters, they will flag up candidate outlier points to allow a person to get involved and make the final determination. This is an important point that needs to be built into any data exploration process.
The simplest operator is
Detect Outlier (Distances). Each example is considered in turn and the distance to the kth nearest neighbor is determined (k is provided as a parameter). ...