immuneML.ml_methods.helper_methods package

Submodules

immuneML.ml_methods.helper_methods.FurthestNeighborClassifier module

class immuneML.ml_methods.helper_methods.FurthestNeighborClassifier.FurthestNeighborClassifier(metric: str = 'precomputed', **kwargs)

Bases: object

Furthest Neighbor Classifier for clustering tasks. It predicts the label (cluster) of an example as the label corresponding to the minimal maximum distance across all labels in training data. The metric used for distance computation can be any metric supported by sklearn.metrics.DistanceMetric, or precomputed (e.g., when using TCRdistEncoder).

fit(X, y)

predict(X)

Module contents