Cost-Sensitive Label Embedding with Multidimensional Scaling¶

class skmultilearn.embedding.CLEMS(measure, is_score=False, params=None)[source]¶

Bases: sklearn.base.BaseEstimator

Embed the label space using a label network embedder from OpenNE

Parameters:

measure (Callable) – a cost function executed on two label vectors dimension (int) – the dimension of the label embedding vectors is_score (boolean) – set to True if measures is a score function (higher value is better), False if loss function (lower is better) param_dict (dict or None) – parameters passed to the embedder, don’t use the dimension and graph parameters, this class will set them at fit

Example code for using this embedder looks like this:

from skmultilearn.embedding import CLEMS, EmbeddingClassifier
from sklearn.ensemble import RandomForestRegressor
from skmultilearn.adapt import MLkNN
from sklearn.metrics import accuracy_score

clf = EmbeddingClassifier(
    CLEMS(accuracy_score, True),
    RandomForestRegressor(n_estimators=10),
    MLkNN(k=5)
)

clf.fit(X_train, y_train)

predictions = clf.predict(X_test)

fit(X, y)[source]¶

Fits the embedder to data

Parameters:	X (array_like, numpy.matrix or scipy.sparse matrix, shape=(n_samples, n_features)) – input feature matrix y (array_like, numpy.matrix or scipy.sparse matrix of {0, 1}, shape=(n_samples, n_labels)) – binary indicator matrix with label assignments
Returns:	fitted instance of self
Return type:	self

fit_transform(X, y)[source]¶

Fit the embedder and transform the output space

Parameters:	X (array_like, numpy.matrix or scipy.sparse matrix, shape=(n_samples, n_features)) – input feature matrix y (array_like, numpy.matrix or scipy.sparse matrix of {0, 1}, shape=(n_samples, n_labels)) – binary indicator matrix with label assignments
Returns:	results of the embedding, input and output space
Return type:	X, y_embedded