class kenchi.outlier_detection.distance_based.KNN(aggregate=False, algorithm='auto', contamination=0.1, leaf_size=30, metric='minkowski', novelty=False, n_jobs=1, n_neighbors=20, p=2, metric_params=None)[source]

Bases: kenchi.outlier_detection.base.BaseOutlierDetector

Outlier detector using k-nearest neighbors algorithm.

Parameters:
  • aggregate (bool, default False) – If True, return the sum of the distances from k nearest neighbors as the anomaly score.
  • algorithm (str, default 'auto') – Tree algorithm to use. Valid algorithms are [‘kd_tree’|’ball_tree’|’auto’].
  • contamination (float, default 0.1) – Proportion of outliers in the data set. Used to define the threshold.
  • leaf_size (int, default 30) – Leaf size of the underlying tree.
  • metric (str or callable, default 'minkowski') – Distance metric to use.
  • novelty (bool, default False) – If True, you can use predict, decision_function and anomaly_score on new unseen data and not on the training data.
  • n_jobs (int, default 1) – Number of jobs to run in parallel. If -1, then the number of jobs is set to the number of CPU cores.
  • n_neighbors (int, default 20) – Number of neighbors.
  • p (int, default 2) – Power parameter for the Minkowski metric.
  • metric_params (dict, default None) – Additioal parameters passed to the requested metric.
anomaly_score_

array-like of shape (n_samples,) – Anomaly score for each training data.

contamination_

float – Actual proportion of outliers in the data set.

threshold_

float – Threshold.

n_neighbors_

int – Actual number of neighbors used for kneighbors queries.

References

[1]Angiulli, F., and Pizzuti, C., “Fast outlier detection in high dimensional spaces,” In Proceedings of PKDD, pp. 15-27, 2002.
[2]Ramaswamy, S., Rastogi, R., and Shim, K., “Efficient algorithms for mining outliers from large data sets,” In Proceedings of SIGMOD, pp. 427-438, 2000.

Examples

>>> import numpy as np
>>> from kenchi.outlier_detection import KNN
>>> X = np.array([
...     [0., 0.], [1., 1.], [2., 0.], [3., -1.], [4., 0.],
...     [5., 1.], [6., 0.], [7., -1.], [8., 0.], [1000., 1.]
... ])
>>> det = KNN(n_neighbors=3)
>>> det.fit_predict(X)
array([ 1,  1,  1,  1,  1,  1,  1,  1,  1, -1])
X_

array-like of shape (n_samples, n_features) – Training data.

class kenchi.outlier_detection.distance_based.OneTimeSampling(contamination=0.1, metric='euclidean', novelty=False, n_subsamples=20, random_state=None, metric_params=None)[source]

Bases: kenchi.outlier_detection.base.BaseOutlierDetector

One-time sampling.

Parameters:
  • contamination (float, default 0.1) – Proportion of outliers in the data set. Used to define the threshold.
  • metric (str, default 'euclidean') – Distance metric to use.
  • novelty (bool, default False) – If True, you can use predict, decision_function and anomaly_score on new unseen data and not on the training data.
  • n_subsamples (int, default 20) – Number of random samples to be used.
  • random_state (int, RandomState instance, default None) – Seed of the pseudo random number generator.
  • metric_params (dict, default None) – Additional parameters passed to the requested metric.
anomaly_score_

array-like of shape (n_samples,) – Anomaly score for each training data.

contamination_

float – Actual proportion of outliers in the data set.

threshold_

float – Threshold.

subsamples_

array-like of shape (n_subsamples,) – Indices of subsamples.

S_

array-like of shape (n_subsamples, n_features) – Subset of the given training data.

References

[3]Sugiyama, M., and Borgwardt, K., “Rapid distance-based outlier detection via sampling,” Advances in NIPS, pp. 467-475, 2013.

Examples

>>> import numpy as np
>>> from kenchi.outlier_detection import OneTimeSampling
>>> X = np.array([
...     [0., 0.], [1., 1.], [2., 0.], [3., -1.], [4., 0.],
...     [5., 1.], [6., 0.], [7., -1.], [8., 0.], [1000., 1.]
... ])
>>> det = OneTimeSampling(n_subsamples=3, random_state=0)
>>> det.fit_predict(X)
array([ 1,  1,  1,  1,  1,  1,  1,  1,  1, -1])