Metric

public abstract class Metric

Known Direct Subclasses

Reduce<T extends TNumber>

Known Indirect Subclasses

BinaryCrossentropy<T extends TNumber>, CategoricalCrossentropy<T extends TNumber>, CategoricalHinge<T extends TNumber>, CosineSimilarity<T extends TNumber>, Hinge<T extends TNumber>, KLDivergence<T extends TNumber>, LogCoshError<T extends TNumber>, Mean<T extends TNumber>, MeanAbsoluteError<T extends TNumber>, MeanAbsolutePercentageError<T extends TNumber>, MeanMetricWrapper<T extends TNumber>, MeanSquaredError<T extends TNumber>, MeanSquaredLogarithmicError<T extends TNumber>, Poisson<T extends TNumber>, SparseCategoricalCrossentropy<T extends TNumber>, SquaredHinge<T extends TNumber>

BinaryCrossentropy<T extends TNumber>	A Metric that computes the binary cross-entropy loss between true labels and predicted labels.
CategoricalCrossentropy<T extends TNumber>	A Metric that computes the categorical cross-entropy loss between true labels and predicted labels.
CategoricalHinge<T extends TNumber>	A Metric that computes the categorical hinge loss metric between labels and predictions.
CosineSimilarity<T extends TNumber>	A metric that computes the cosine similarity metric between labels and predictions.
Hinge<T extends TNumber>	A metric that computes the hinge loss metric between labels and predictions.
KLDivergence<T extends TNumber>	A metric that computes the Kullback-Leibler divergence loss metric between labels and predictions.
LogCoshError<T extends TNumber>	A metric that computes the logarithm of the hyperbolic cosine of the prediction error metric between labels and predictions.
Mean<T extends TNumber>	A metric that that implements a weighted mean `WEIGHTED_MEAN`
MeanAbsoluteError<T extends TNumber>	A metric that computes the mean of absolute difference between labels and predictions.
MeanAbsolutePercentageError<T extends TNumber>	A metric that computes the mean of absolute difference between labels and predictions.
MeanMetricWrapper<T extends TNumber>	A class that bridges a stateless loss function with the `Mean` metric using a reduction of `WEIGHTED_MEAN`.
MeanSquaredError<T extends TNumber>	A metric that computes the mean of absolute difference between labels and predictions.
MeanSquaredLogarithmicError<T extends TNumber>	A metric that computes the mean of absolute difference between labels and predictions.
Poisson<T extends TNumber>	A metric that computes the poisson loss metric between labels and predictions.
SparseCategoricalCrossentropy<T extends TNumber>	A metric that computes the sparse categorical cross-entropy loss between true labels and predicted labels.
SquaredHinge<T extends TNumber>	A metric that computes the squared hinge loss metric between labels and predictions.

Base class for Metrics

Public Methods

final Operand<T>	callOnce(Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Calls update state once, followed by a call to get the result
String	getName() Gets the name of this metric.
long	getSeed() Gets the random number generator seed value
Ops	getTF() Gets the TensorFlow Ops
abstract Op	resetStates() Resets any state variables to their initial values
abstract Operand<T>	result() Gets the current result of the metric
final Op	updateState(Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a NoOp Operation with control dependencies to update the metric state
final Op	updateState(Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Creates a NoOp Operation with control dependencies to update the metric state
List<Op>	updateStateList(Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on labels and predictions.
List<Op>	updateStateList(Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights) Creates a List of Operations to update the metric state based on input values.

Inherited Methods

From class java.lang.Object

boolean	equals(Object arg0)
final Class<?>	getClass()
int	hashCode()
final void	notify()
final void	notifyAll()
String	toString()
final void	wait(long arg0, int arg1)
final void	wait(long arg0)
final void	wait()

Public Methods

public final Operand<T> callOnce (Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights)

Calls update state once, followed by a call to get the result

Parameters

values	the inputs to be passed to update state, this may not be null
sampleWeights	sample weights to be applied to values, may be null.

Returns

the result, possibly with control dependencies

public String getName ()

Gets the name of this metric.

Returns

the name of this metric

public long getSeed ()

Gets the random number generator seed value

Returns

the random number generator seed value

public Ops getTF ()

Gets the TensorFlow Ops

Returns

the TensorFlow Ops

public abstract Op resetStates ()

Resets any state variables to their initial values

Returns

the control operation for doing the reset

public abstract Operand<T> result ()

Gets the current result of the metric

Returns

the result, possibly with control dependencies

public final Op updateState (Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights)

Creates a NoOp Operation with control dependencies to update the metric state

Parameters

labels	the labels
predictions	the predictions
sampleWeights	sample weights to be applied to values, may be null.

Returns

the Operation to update the metric state

public final Op updateState (Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights)

Creates a NoOp Operation with control dependencies to update the metric state

Parameters

values	the inputs to be passed to update state, this may not be null
sampleWeights	sample weights to be applied to values, may be null.

Returns

the Operation to update the metric state

public List<Op> updateStateList (Operand<? extends TNumber> labels, Operand<? extends TNumber> predictions, Operand<? extends TNumber> sampleWeights)

Creates a List of Operations to update the metric state based on labels and predictions.

This is an empty implementation that should be overridden in a sub class, if needed.

Parameters

labels	the labels
predictions	the predictions
sampleWeights	sample weights to be applied to values, may be null.

Returns

a List of Operations to update the metric state

public List<Op> updateStateList (Operand<? extends TNumber> values, Operand<? extends TNumber> sampleWeights)

Creates a List of Operations to update the metric state based on input values.

This is an empty implementation that should be overridden in a subclass, if needed.

Parameters

values	the inputs to be passed to update state, this may not be null
sampleWeights	sample weights to be applied to values, may be null.

Returns

a List of Operations to update the metric state