pusion.core.simple_average_combiner

class pusion.core.simple_average_combiner.SimpleAverageCombiner

Bases: pusion.core.combiner.UtilityBasedCombiner

The SimpleAverageCombiner (AVG) fuses decisions using the arithmetic mean rule. The mean is calculated between decision vectors obtained by multiple ensemble classifiers for a sample. The AVG combiner is unaware of the input problem (multiclass/multilabel) or the assignment type (crisp/continuous).

combine(decision_tensor)

Combine decision outputs by averaging the class support of each classifier in the given ensemble.

Parameters

decision_tensor – numpy.array of shape (n_classifiers, n_samples, n_classes). Tensor of either crisp or continuous decision outputs by different classifiers per sample.

Returns

A matrix (numpy.array) of crisp assignments which represents fused decisions obtained by the AVG method. Axis 0 represents samples and axis 1 the class assignments which are aligned with axis 2 in decision_tensor input tensor.

class pusion.core.simple_average_combiner.CRSimpleAverageCombiner

Bases: pusion.core.simple_average_combiner.SimpleAverageCombiner

The CRSimpleAverageCombiner is a modification of SimpleAverageCombiner that also supports complementary-redundant decision outputs. Therefore the input is transformed to a unified tensor representation supporting undefined class assignments. The mean is calculated only for assignments which are defined. To call combine() a coverage needs to be set first by the inherited set_coverage() method.

combine(decision_outputs)

Combine decision outputs by averaging the defined class support of each classifier in the given ensemble. Undefined class supports are excluded from averaging.

Parameters

decision_outputs – list of numpy.array matrices, each of shape (n_samples, n_classes’), where n_classes’ is classifier-specific and described by the coverage. Each matrix corresponds to one of n_classifiers classifiers and contains either crisp or continuous decision outputs per sample.

Returns

A matrix (numpy.array) of crisp assignments which represents fused decisions obtained by the AVG method. Axis 0 represents samples and axis 1 the class assignments which are aligned with axis 2 in decision_tensor input tensor.