pusion.auto.detector module

pusion.auto.detector.determine_problem(decision_outputs)

Determine the classification problem based on the decision outputs.

Parameters

decision_outputsnumpy.array of shape (n_classifiers, n_samples, n_classes) or a list of numpy.array elements of shape (n_samples, n_classes’), where n_classes’ is classifier-specific due to the coverage.

Returns

string constant ‘MULTI_CLASS’ or ‘MULTI_LABEL’. See pusion.util.constants.Problem.

pusion.auto.detector.determine_assignment_type(decision_outputs)

Determine the assignment type based on the decision outputs.

Parameters

decision_outputsnumpy.array of shape (n_classifiers, n_samples, n_classes) or a list of numpy.array elements of shape (n_samples, n_classes’), where n_classes’ is classifier-specific due to the coverage.

Returns

string constant ‘CRISP’ or ‘CONTINUOUS’. See pusion.util.constants.AssignmentType.

pusion.auto.detector.determine_coverage_type(coverage)

Determine the coverage type.

Parameters

coveragelist of list elements. Each inner list contains classes as integers covered by a classifier, which is identified by the positional index of the respective list.

Returns

string constant ‘REDUNDANT’, ‘COMPLEMENTARY’ or ‘COMPLEMENTARY_REDUNDANT’. See pusion.util.constants.CoverageType.

pusion.auto.detector.determine_pac(decision_outputs, coverage=None)

Determine the PAC-tuple (problem, assignment type and coverage type) based on the given decision outputs and coverage.

Parameters
  • decision_outputsnumpy.array of shape (n_classifiers, n_samples, n_classes) or a list of numpy.array elements of shape (n_samples, n_classes’), where n_classes’ is classifier-specific due to the coverage.

  • coveragelist of list elements. Each inner list contains classes as integers covered by a classifier, which is identified by the positional index of the respective list.

Returns

tuple of string constants representing the PAC. See pusion.util.constants.Problem, pusion.util.constants.AssignmentType and pusion.util.constants.CoverageType.