milia.anomaly.models#
Registered anomaly detectors, one module per model (see milia.core.registry).
Modules
Shewhart control limit: the largest standardised deviation of any reading from its training mean. |
|
COPOD: a skew-weighted empirical copula. |
|
ECOD: empirical-CDF tail probabilities, summed over dimensions. |
|
HBOS: histogram-based outlier score. |
|
Isolation Forest: how few random splits it takes to isolate a point. |
|
Kernel density estimate: how far into a low-density region a point sits. |
|
k-nearest-neighbour distance to the training data. |
|
Local Outlier Factor: local density relative to a point's own neighbours. |
|
Robust Mahalanobis distance from a Minimum Covariance Determinant fit. |
|
One-class SVM: distance from a learned support boundary. |
|
Reconstruction error from the leading principal components (the Q statistic of multivariate process control). |