org.tribuo.anomaly.evaluation

Interface AnomalyEvaluation

All Superinterfaces:

Evaluation<Event>, com.oracle.labs.mlrg.olcut.provenance.Provenancable<EvaluationProvenance>

public interface AnomalyEvaluation
extends Evaluation<Event>

An Evaluation for anomaly detection Events.

Method Summary

Modifier and Type	Method and Description
String	confusionString() Returns a confusion matrix formatted String for display.
double	getF1() Returns the F_1 score of the anomalous events, i.e., the harmonic mean of the precision and the recall.
long	getFalseNegatives() Returns the number of false negatives, i.e., anomalous events classified as expected.
long	getFalsePositives() Returns the number of false positives, i.e., expected events classified as anomalous.
double	getPrecision() Returns the precision of the anomalous events, i.e., true positives divided by the number of predicted positives.
double	getRecall() Returns the recall of the anomalous events, i.e., true positives divided by the number of positives.
long	getTrueNegatives() Returns the number of true negatives, i.e., expected events classified as events.
long	getTruePositives() Returns the number of true positives, i.e., anomalous events classified as anomalous.

Methods inherited from interface org.tribuo.evaluation.Evaluation

asMap, get, getPredictions

Methods inherited from interface com.oracle.labs.mlrg.olcut.provenance.Provenancable

getProvenance

Method Detail

getFalsePositives

long getFalsePositives()

Returns the number of false positives, i.e., expected events classified as anomalous.

Returns:

The number of false positives.

getTruePositives

long getTruePositives()

Returns the number of true positives, i.e., anomalous events classified as anomalous.

Returns:

The number of true positives.

getTrueNegatives

long getTrueNegatives()

Returns the number of true negatives, i.e., expected events classified as events.

Returns:

The number of true negatives.

getFalseNegatives

long getFalseNegatives()

Returns the number of false negatives, i.e., anomalous events classified as expected.

These are the ones you don't want.

Returns:

The number of false negatives.

getPrecision

double getPrecision()

Returns the precision of the anomalous events, i.e., true positives divided by the number of predicted positives.

Returns:

The precision.

getRecall

double getRecall()

Returns the recall of the anomalous events, i.e., true positives divided by the number of positives.

Returns:

The recall.

getF1

double getF1()

Returns the F_1 score of the anomalous events, i.e., the harmonic mean of the precision and the recall.

Returns:

The F_1 score.

confusionString

String confusionString()

Returns a confusion matrix formatted String for display.

Returns:

The confusion matrix in a String.