skclean.detectors.ForestKDN

class skclean.detectors.ForestKDN(n_neighbors=5, n_estimators=100, max_leaf_nodes=64, weight='distance', n_jobs=1, random_state=None)

Like KDN, but a trained Random Forest is used to compute pairwise similarity.

Specifically, for a pair of samples, their similarity is the percentage of times they belong to the same leaf. See [LM17] for details.

Parameters
  • n_neighbors (int, default=5) – No of nearest neighbors to use to compute conf_score

  • n_estimators (int, default=101) – No of trees in Random Forest.

  • max_leaf_nodes (int, default=64) – Maximum no of leaves in each tree.

  • weight (string, default='distance') – weight function used in prediction. If ‘distance’, weights points by the inverse of their distance. If ‘uniform’, all points in each neighborhood are weighted equally.

  • n_jobs (int, default=1) – No of parallel cpu cores to use

  • random_state (int, default=None) – Set this value for reproducibility

Methods

__init__([n_neighbors, n_estimators, …])

Initialize self.

detect(X, y)

fit_transform(X[, y])

Fit to data, then transform it.

get_params([deep])

Get parameters for this estimator.

set_params(**params)

Set the parameters of this estimator.

transform(X)