Class ClassifierEvaluator<TInstanceSource, TInstance, TLabelSource, TLabel>
Evaluates the predictions of a classifier.
Inherited Members
Namespace: Microsoft.ML.Probabilistic.Learners
Assembly: Microsoft.ML.Probabilistic.Learners.Classifier.dll
Syntax
public class ClassifierEvaluator<TInstanceSource, TInstance, TLabelSource, TLabel>Type Parameters
| Name | Description |
|---|---|
| TInstanceSource | The type of a source of instances. |
| TInstance | The type of an instance. |
| TLabelSource | The type of a source of labels. |
| TLabel | The type of a label. |
Remarks
Assumes that there are as many predictions as ground truth instances and that the order of the predictions matches the order of the ground truth instances.
Constructors
ClassifierEvaluator(IClassifierEvaluatorMapping<TInstanceSource, TInstance, TLabelSource, TLabel>)
Initializes a new instance of the ClassifierEvaluator<TInstanceSource, TInstance, TLabelSource, TLabel> class.
Declaration
public ClassifierEvaluator(IClassifierEvaluatorMapping<TInstanceSource, TInstance, TLabelSource, TLabel> mapping)Parameters
| Type | Name | Description |
|---|---|---|
| IClassifierEvaluatorMapping<TInstanceSource, TInstance, TLabelSource, TLabel> | mapping | The mapping to access the ground truth labels. |
Methods
AreaUnderRocCurve(TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance or label source.
See Hand, D.J. and Till, R.J. (2001): "A simple generalization of the Area Under the ROC Curve for Multiple Class Classification Problems". Machine Learning, 45, pages 171-186.
Declaration
public double AreaUnderRocCurve(TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. Returns NaN if there are no positive or no negative instances. |
AreaUnderRocCurve(TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>, out IDictionary<TLabel, IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance or label source.
See Hand, D.J. and Till, R.J. (2001): "A simple generalization of the Area Under the ROC Curve for Multiple Class Classification Problems". Machine Learning, 45, pages 171-186.
Declaration
public double AreaUnderRocCurve(TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions, out IDictionary<TLabel, IDictionary<TLabel, double>> aucMatrix)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| IDictionary<TLabel, IDictionary<TLabel, Double>> | aucMatrix | A matrix of pairwise AUC metrics. If the area under the curve is not defined for a given pair of classes, the corresponding element in the matrix is NaN. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. Returns NaN if there are no positive or no negative instances. |
AreaUnderRocCurve(TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance source.
See Hand, D.J. and Till, R.J. (2001): "A simple generalization of the Area Under the ROC Curve for Multiple Class Classification Problems". Machine Learning, 45, pages 171-186.
Declaration
public double AreaUnderRocCurve(TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. Returns NaN if there are no positive or no negative instances. |
AreaUnderRocCurve(TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>, out IDictionary<TLabel, IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance source.
See Hand, D.J. and Till, R.J. (2001): "A simple generalization of the Area Under the ROC Curve for Multiple Class Classification Problems". Machine Learning, 45, pages 171-186.
Declaration
public double AreaUnderRocCurve(TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions, out IDictionary<TLabel, IDictionary<TLabel, double>> aucMatrix)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| IDictionary<TLabel, IDictionary<TLabel, Double>> | aucMatrix | A matrix of pairwise AUC metrics. If the area under the curve is not defined for a given pair of classes, the corresponding element in the matrix is NaN. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. Returns NaN if there are no positive or no negative instances. |
AreaUnderRocCurve(TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance or label source.
Declaration
public double AreaUnderRocCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. |
AreaUnderRocCurve(TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance source.
Declaration
public double AreaUnderRocCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. |
AreaUnderRocCurve(TLabel, TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance or label source.
Declaration
public double AreaUnderRocCurve(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. |
AreaUnderRocCurve(TLabel, TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the area under the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance source.
Declaration
public double AreaUnderRocCurve(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| Double | The area under the receiver operating characteristic curve. |
CalibrationCurve(TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the empirical probability calibration curve for the class of the specified label.
Declaration
public IEnumerable<CalibrationPair> CalibrationCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the class to compute the curve for. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<CalibrationPair> | The computed empirical probability calibration curve. |
CalibrationCurve(TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>, Int32, Int32)
Computes the empirical probability calibration curve for the class of the specified label.
Declaration
public IEnumerable<CalibrationPair> CalibrationCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions, int binCount, int minBinInstanceCount = 1)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the class to compute the curve for. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| Int32 | binCount | The number of bins to use. |
| Int32 | minBinInstanceCount | The minimal number of instances per bin. Defaults to 1. |
Returns
| Type | Description |
|---|---|
| IEnumerable<CalibrationPair> | The computed empirical probability calibration curve. |
CalibrationCurve(TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the empirical probability calibration curve for the class of the specified label.
Declaration
public IEnumerable<CalibrationPair> CalibrationCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the class to compute the curve for. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<CalibrationPair> | The computed empirical probability calibration curve. |
CalibrationCurve(TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>, Int32, Int32)
Computes the empirical probability calibration curve for the class of the specified label.
Declaration
public IEnumerable<CalibrationPair> CalibrationCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions, int binCount, int minBinInstanceCount = 1)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the class to compute the curve for. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| Int32 | binCount | The number of bins to use. |
| Int32 | minBinInstanceCount | The minimal number of instances per bin. Defaults to 1. |
Returns
| Type | Description |
|---|---|
| IEnumerable<CalibrationPair> | The computed empirical probability calibration curve. |
ConfusionMatrix(TInstanceSource, TLabelSource, IEnumerable<TLabel>)
Computes a confusion matrix for specified predictions given the ground truth provided by an instance or label source.
Declaration
public ConfusionMatrix<TLabel> ConfusionMatrix(TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<TLabel> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<TLabel> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| ConfusionMatrix<TLabel> | The computed confusion matrix. |
ConfusionMatrix(TInstanceSource, IEnumerable<TLabel>)
Computes a confusion matrix for specified predictions given the ground truth provided by an instance source.
Declaration
public ConfusionMatrix<TLabel> ConfusionMatrix(TInstanceSource instanceSource, IEnumerable<TLabel> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<TLabel> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| ConfusionMatrix<TLabel> | The computed confusion matrix. |
Evaluate<TPrediction>(TInstanceSource, TLabelSource, IEnumerable<TPrediction>, Func<TLabel, TPrediction, Double>)
Computes a performance metric for specified predictions and ground truth as provided by a given instance or label source.
Declaration
public double Evaluate<TPrediction>(TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<TPrediction> predictions, Func<TLabel, TPrediction, double> performanceMetric)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<TPrediction> | predictions | The predictions. |
| Func<TLabel, TPrediction, Double> | performanceMetric | The performance metric. |
Returns
| Type | Description |
|---|---|
| Double | The computed performance metric. |
Type Parameters
| Name | Description |
|---|---|
| TPrediction | The type of the prediction. |
Evaluate<TPrediction>(TInstanceSource, IEnumerable<TPrediction>, Func<TLabel, TPrediction, Double>)
Computes a performance metric for specified predictions and ground truth as provided by a given instance source.
Declaration
public double Evaluate<TPrediction>(TInstanceSource instanceSource, IEnumerable<TPrediction> predictions, Func<TLabel, TPrediction, double> performanceMetric)Parameters
| Type | Name | Description |
|---|---|---|
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<TPrediction> | predictions | The predictions. |
| Func<TLabel, TPrediction, Double> | performanceMetric | The performance metric. |
Returns
| Type | Description |
|---|---|
| Double | The computed performance metric. |
Type Parameters
| Name | Description |
|---|---|
| TPrediction | The type of the prediction. |
EvaluateOneVersusAnother<TResult>(TLabel, TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>, Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult>)
Computes a performance metric for specified positive and negative classes.
Declaration
public TResult EvaluateOneVersusAnother<TResult>(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions, Func<IEnumerable<int>, IDictionary<int, double>, TResult> performanceMetric)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult> | performanceMetric | The performance metric. |
Returns
| Type | Description |
|---|---|
| TResult | The computed performance metric. |
Type Parameters
| Name | Description |
|---|---|
| TResult | The result of the performance metric. |
EvaluateOneVersusAnother<TResult>(TLabel, TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>, Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult>)
Computes a performance metric for specified positive and negative classes.
Declaration
public TResult EvaluateOneVersusAnother<TResult>(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions, Func<IEnumerable<int>, IDictionary<int, double>, TResult> performanceMetric)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult> | performanceMetric | The performance metric. |
Returns
| Type | Description |
|---|---|
| TResult | The computed performance metric. |
Type Parameters
| Name | Description |
|---|---|
| TResult | The result of the performance metric. |
EvaluateOneVersusRest<TResult>(TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>, Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult>)
Computes a performance metric for a specified positive class, assuming all other classes are negative.
Declaration
public TResult EvaluateOneVersusRest<TResult>(TLabel positiveClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions, Func<IEnumerable<int>, IDictionary<int, double>, TResult> performanceMetric)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult> | performanceMetric | The performance metric. |
Returns
| Type | Description |
|---|---|
| TResult | The computed performance metric. |
Type Parameters
| Name | Description |
|---|---|
| TResult | The result of the performance metric. |
EvaluateOneVersusRest<TResult>(TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>, Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult>)
Computes a performance metric for a specified positive class, assuming all other classes are negative.
Declaration
public TResult EvaluateOneVersusRest<TResult>(TLabel positiveClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions, Func<IEnumerable<int>, IDictionary<int, double>, TResult> performanceMetric)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
| Func<IEnumerable<Int32>, IDictionary<Int32, Double>, TResult> | performanceMetric | The performance metric. |
Returns
| Type | Description |
|---|---|
| TResult | The computed performance metric. |
Type Parameters
| Name | Description |
|---|---|
| TResult | The result of the performance metric. |
PrecisionRecallCurve(TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the precision-recall curve for specified predictions given the ground truth provided by an instance or label source.
Declaration
public IEnumerable<PrecisionRecall> PrecisionRecallCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<PrecisionRecall> | The computed precision-recall curve. |
PrecisionRecallCurve(TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the precision-recall curve for specified predictions given the ground truth provided by an instance source.
Declaration
public IEnumerable<PrecisionRecall> PrecisionRecallCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<PrecisionRecall> | The computed precision-recall curve. |
PrecisionRecallCurve(TLabel, TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the precision-recall curve for specified predictions given the ground truth provided by an instance or label source.
Declaration
public IEnumerable<PrecisionRecall> PrecisionRecallCurve(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<PrecisionRecall> | The computed precision-recall curve. |
PrecisionRecallCurve(TLabel, TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the precision-recall curve for specified predictions given the ground truth provided by an instance source.
Declaration
public IEnumerable<PrecisionRecall> PrecisionRecallCurve(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<PrecisionRecall> | The computed precision-recall curve. |
ReceiverOperatingCharacteristicCurve(TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance or label source.
Declaration
public IEnumerable<FalseAndTruePositiveRate> ReceiverOperatingCharacteristicCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<FalseAndTruePositiveRate> | The computed receiver operating characteristic curve. |
ReceiverOperatingCharacteristicCurve(TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance source.
Declaration
public IEnumerable<FalseAndTruePositiveRate> ReceiverOperatingCharacteristicCurve(TLabel positiveClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<FalseAndTruePositiveRate> | The computed receiver operating characteristic curve. |
ReceiverOperatingCharacteristicCurve(TLabel, TLabel, TInstanceSource, TLabelSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance or label source.
Declaration
public IEnumerable<FalseAndTruePositiveRate> ReceiverOperatingCharacteristicCurve(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, TLabelSource labelSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| TLabelSource | labelSource | The label source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<FalseAndTruePositiveRate> | The computed receiver operating characteristic curve. |
ReceiverOperatingCharacteristicCurve(TLabel, TLabel, TInstanceSource, IEnumerable<IDictionary<TLabel, Double>>)
Computes the receiver operating characteristic curve for specified predictions given the ground truth provided by an instance source.
Declaration
public IEnumerable<FalseAndTruePositiveRate> ReceiverOperatingCharacteristicCurve(TLabel positiveClassLabel, TLabel negativeClassLabel, TInstanceSource instanceSource, IEnumerable<IDictionary<TLabel, double>> predictions)Parameters
| Type | Name | Description |
|---|---|---|
| TLabel | positiveClassLabel | The label of the positive class. |
| TLabel | negativeClassLabel | The label of the negative class. |
| TInstanceSource | instanceSource | The instance source. |
| IEnumerable<IDictionary<TLabel, Double>> | predictions | The predictions. |
Returns
| Type | Description |
|---|---|
| IEnumerable<FalseAndTruePositiveRate> | The computed receiver operating characteristic curve. |