- class AllPairs¶
The All-Pairs multiclass extension.
In the all-pairs reduction, one trains \(k(k−1)/2\) binary classifiers for a \(k\)-way multiclass problem; each receives the samples of a pair of classes from the original training set, and must learn to distinguish these two classes. At prediction time, a weighted voting scheme is used: all \(k(k−1)/2\) classifiers are applied to an unseen sample, and each class gets assigned the sum of all the scores obtained by the various classifiers. The combined classifier returns as a result the class getting the highest value.
Applying multiple estimators for prediction.
Called internally to set
Estimatorand parameters :type estimator_cls:
Estimator] :param estimator_cls: An
Estimatorclass :type params:
List] :param params: Parameters for the estimator
Testing multiple estimators each for distinguishing a pair of classes.
Training multiple estimators each for distinguishing a pair of classes.