class MulticlassMetrics extends AnyRef
Evaluator for multiclass classification.
- Annotations
- @Since( "1.1.0" )
- Source
- MulticlassMetrics.scala
- Alphabetic
- By Inheritance
- MulticlassMetrics
- AnyRef
- Any
- Hide All
- Show All
- Public
- All
Instance Constructors
Value Members
-
lazy val
accuracy: Double
Returns accuracy (equals to the total number of correctly classified instances out of the total number of instances.)
Returns accuracy (equals to the total number of correctly classified instances out of the total number of instances.)
- Annotations
- @Since( "2.0.0" )
-
def
confusionMatrix: Matrix
Returns confusion matrix: predicted classes are in columns, they are ordered by class label ascending, as in "labels"
Returns confusion matrix: predicted classes are in columns, they are ordered by class label ascending, as in "labels"
- Annotations
- @Since( "1.1.0" )
-
def
fMeasure(label: Double): Double
Returns f1-measure for a given label (category)
Returns f1-measure for a given label (category)
- label
the label.
- Annotations
- @Since( "1.1.0" )
-
def
fMeasure(label: Double, beta: Double): Double
Returns f-measure for a given label (category)
Returns f-measure for a given label (category)
- label
the label.
- beta
the beta parameter.
- Annotations
- @Since( "1.1.0" )
-
def
falsePositiveRate(label: Double): Double
Returns false positive rate for a given label (category)
Returns false positive rate for a given label (category)
- label
the label.
- Annotations
- @Since( "1.1.0" )
-
lazy val
hammingLoss: Double
Returns Hamming-loss
Returns Hamming-loss
- Annotations
- @Since( "3.0.0" )
-
lazy val
labels: Array[Double]
Returns the sequence of labels in ascending order
Returns the sequence of labels in ascending order
- Annotations
- @Since( "1.1.0" )
-
def
logLoss(eps: Double = 1e-15): Double
Returns the log-loss, aka logistic loss or cross-entropy loss.
Returns the log-loss, aka logistic loss or cross-entropy loss.
- eps
log-loss is undefined for p=0 or p=1, so probabilities are clipped to max(eps, min(1 - eps, p)).
- Annotations
- @Since( "3.0.0" )
-
def
precision(label: Double): Double
Returns precision for a given label (category)
Returns precision for a given label (category)
- label
the label.
- Annotations
- @Since( "1.1.0" )
-
def
recall(label: Double): Double
Returns recall for a given label (category)
Returns recall for a given label (category)
- label
the label.
- Annotations
- @Since( "1.1.0" )
-
def
truePositiveRate(label: Double): Double
Returns true positive rate for a given label (category)
Returns true positive rate for a given label (category)
- label
the label.
- Annotations
- @Since( "1.1.0" )
-
def
weightedFMeasure(beta: Double): Double
Returns weighted averaged f-measure
Returns weighted averaged f-measure
- beta
the beta parameter.
- Annotations
- @Since( "1.1.0" )
-
lazy val
weightedFMeasure: Double
Returns weighted averaged f1-measure
Returns weighted averaged f1-measure
- Annotations
- @Since( "1.1.0" )
-
lazy val
weightedFalsePositiveRate: Double
Returns weighted false positive rate
Returns weighted false positive rate
- Annotations
- @Since( "1.1.0" )
-
lazy val
weightedPrecision: Double
Returns weighted averaged precision
Returns weighted averaged precision
- Annotations
- @Since( "1.1.0" )
-
lazy val
weightedRecall: Double
Returns weighted averaged recall (equals to precision, recall and f-measure)
Returns weighted averaged recall (equals to precision, recall and f-measure)
- Annotations
- @Since( "1.1.0" )
-
lazy val
weightedTruePositiveRate: Double
Returns weighted true positive rate (equals to precision, recall and f-measure)
Returns weighted true positive rate (equals to precision, recall and f-measure)
- Annotations
- @Since( "1.1.0" )