Package org.tribuo.classification.evaluation

Evaluation classes for multi-class classification.

See: Description

Interface Summary

Interface	Description
ClassifierEvaluation<T extends Classifiable<T>>	Defines methods that calculate classification performance, used for both multi-class and multi-label classification.
ConfusionMatrix<T extends Classifiable<T>>	A confusion matrix for Classifiables.
LabelEvaluation	Adds multi-class classification specific metrics to ClassifierEvaluation.

Class Summary

Class	Description
ConfusionMetrics	Static functions for computing classification metrics based on a ConfusionMatrix.
LabelConfusionMatrix	A confusion matrix for Labels.
LabelEvaluationUtil	Static utility functions for calculating performance metrics on Labels.
LabelEvaluationUtil.PRCurve	Stores the Precision-Recall curve as three arrays: the precisions, the recalls, and the thresholds associated with those values.
LabelEvaluationUtil.ROC	Stores the ROC curve as three arrays: the false positive rate, the true positive rate, and the thresholds associated with those rates.
LabelEvaluator	An Evaluator for Labels.
LabelMetric	A EvaluationMetric for Labels which calculates the value based on a ConfusionMatrix.
LabelMetric.Context	The context for a LabelMetric is a ConfusionMatrix.

Enum Summary

Enum	Description
LabelMetrics	An enum of the default LabelMetrics supported by the multi-class classification evaluation package.

Package org.tribuo.classification.evaluation Description

Evaluation classes for multi-class classification. Also used as the basis for the multi-label classification evaluation package.

The default metrics calculated are found in LabelMetrics, and are based on statistics calculated from a ConfusionMatrix.

User specified metrics are not currently supported.