Package ai.djl.training.evaluator

Class BinaryAccuracy

java.lang.Object

ai.djl.training.evaluator.Evaluator

ai.djl.training.evaluator.AbstractAccuracy

ai.djl.training.evaluator.BinaryAccuracy

public class BinaryAccuracy
extends AbstractAccuracy

BinaryAccuracy is the AbstractAccuracy with two classes.

It is assumed that the classes are identified with a labels array of 0s and 1s and a prediction array where values above the threshold are the positive (1) examples and values below the threshold are the negative (0) examples. If you have a different encoding, you may want to look at the Accuracy.

Field Summary

Fields inherited from class ai.djl.training.evaluator.AbstractAccuracy

axis, correctInstances

Fields inherited from class ai.djl.training.evaluator.Evaluator

totalInstances

Constructor Summary

Constructors

Constructor	Description
BinaryAccuracy()	Creates a binary (two class) accuracy evaluator with 0 threshold.
BinaryAccuracy(float threshold)	Creates a binary (two class) accuracy evaluator that computes accuracy across axis 1 along the 0th index.
BinaryAccuracy(String name, float threshold)	Creates a binary (two class) accuracy evaluator that computes accuracy across axis 1 along given index.
BinaryAccuracy(String name, float threshold, int axis)	Creates a binary (two class) accuracy evaluator.

Method Summary

Modifier and Type	Method	Description
protected ai.djl.util.Pair<Long,NDArray>	accuracyHelper(NDList labels, NDList predictions)	A helper for classes extending AbstractAccuracy.

Methods inherited from class ai.djl.training.evaluator.AbstractAccuracy

addAccumulator, evaluate, getAccumulator, resetAccumulator, updateAccumulator, updateAccumulators

Methods inherited from class ai.djl.training.evaluator.Evaluator

checkLabelShapes, checkLabelShapes, getName

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Details

BinaryAccuracy

public BinaryAccuracy(String name,
 float threshold,
 int axis)

Creates a binary (two class) accuracy evaluator.

Parameters:

name - the name of the evaluator, default is "Accuracy"

threshold - the value differentiating the posive and negative classes (usually 0 or .5)

axis - the axis that represent classes in prediction, default 1

BinaryAccuracy

public BinaryAccuracy(String name,
 float threshold)

Creates a binary (two class) accuracy evaluator that computes accuracy across axis 1 along given index.

Parameters:

name - the name of the evaluator, default is "Accuracy"

threshold - the value differentiating the posive and negative classes (usually 0 or .5)

BinaryAccuracy

public BinaryAccuracy(float threshold)

Creates a binary (two class) accuracy evaluator that computes accuracy across axis 1 along the 0th index.

Parameters:

threshold - the value differentiating the posive and negative classes (usually 0 or .5)

BinaryAccuracy

public BinaryAccuracy()

Creates a binary (two class) accuracy evaluator with 0 threshold.

Method Details

accuracyHelper

protected ai.djl.util.Pair<Long,NDArray> accuracyHelper(NDList labels,
 NDList predictions)

A helper for classes extending AbstractAccuracy.

Specified by:

accuracyHelper in class AbstractAccuracy

Parameters:

labels - the labels to get accuracy for

predictions - the predictions to get accuracy for

Returns:

a pair(number of total values, ndarray int of correct values)