drift#
Drift detectors identify if the statistical properties of the data has changed.
Detector Classes#
|
Drift detector employing the Cramér-von Mises (CVM) Drift Detection test. |
|
Drift detector employing the Kolmogorov-Smirnov (KS) distribution test. |
|
Maximum Mean Discrepancy (MMD) Drift Detection algorithm using a permutation test. |
|
Test for a change in the number of instances falling into regions on which the model is uncertain. |
Supporting Classes#
Kernels#
Kernels are used to map non-linear data to a higher dimensional space.
|
Gaussian RBF kernel: k(x,y) = exp(-(1/(2*sigma^2)||x-y||^2). |
Update Strategies#
Update strategies inform how the drift detector classes update the reference data when monitoring for drift.
Output Classes#
|
Output class for |
|