org.deeplearning4j.earlystopping.scorecalc

Class VAEReconProbScoreCalculator

java.lang.Object

org.deeplearning4j.earlystopping.scorecalc.base.BaseScoreCalculator<Model>

org.deeplearning4j.earlystopping.scorecalc.VAEReconProbScoreCalculator

All Implemented Interfaces:

Serializable, ScoreCalculator<Model>

public class VAEReconProbScoreCalculator
extends BaseScoreCalculator<Model>

Score calculator for variational autoencoder reconstruction probability or reconstruction log probability for a MultiLayerNetwork or ComputationGraph. VariationalAutoencoder layer must be first layer in the network
See VariationalAutoencoder.reconstructionProbability(INDArray, int) for more details

Author:

Alex Black

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
protected boolean	average
protected boolean	logProb
protected int	reconstructionProbNumSamples

Fields inherited from class org.deeplearning4j.earlystopping.scorecalc.base.BaseScoreCalculator

exampleCount, iterator, mdsIterator, minibatchCount, scoreSum

Constructor Summary

Constructors

Constructor and Description
VAEReconProbScoreCalculator(org.nd4j.linalg.dataset.api.iterator.DataSetIterator iterator, int reconstructionProbNumSamples, boolean logProb) Constructor for average reconstruction probability
VAEReconProbScoreCalculator(org.nd4j.linalg.dataset.api.iterator.DataSetIterator iterator, int reconstructionProbNumSamples, boolean logProb, boolean average) Constructor for reconstruction probability

Method Summary

Modifier and Type	Method and Description
protected double	finalScore(double scoreSum, int minibatchCount, int exampleCount)
protected org.nd4j.linalg.api.ndarray.INDArray[]	output(Model network, org.nd4j.linalg.api.ndarray.INDArray[] input, org.nd4j.linalg.api.ndarray.INDArray[] fMask, org.nd4j.linalg.api.ndarray.INDArray[] lMask)
protected org.nd4j.linalg.api.ndarray.INDArray	output(Model network, org.nd4j.linalg.api.ndarray.INDArray input, org.nd4j.linalg.api.ndarray.INDArray fMask, org.nd4j.linalg.api.ndarray.INDArray lMask)
protected void	reset()
protected double	scoreMinibatch(Model network, org.nd4j.linalg.api.ndarray.INDArray[] features, org.nd4j.linalg.api.ndarray.INDArray[] labels, org.nd4j.linalg.api.ndarray.INDArray[] fMask, org.nd4j.linalg.api.ndarray.INDArray[] lMask, org.nd4j.linalg.api.ndarray.INDArray[] output)
protected double	scoreMinibatch(Model net, org.nd4j.linalg.api.ndarray.INDArray features, org.nd4j.linalg.api.ndarray.INDArray labels, org.nd4j.linalg.api.ndarray.INDArray fMask, org.nd4j.linalg.api.ndarray.INDArray lMask, org.nd4j.linalg.api.ndarray.INDArray output)

Methods inherited from class org.deeplearning4j.earlystopping.scorecalc.base.BaseScoreCalculator

arr, calculateScore, get0

Methods inherited from class java.lang.Object

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

Field Detail

reconstructionProbNumSamples

protected final int reconstructionProbNumSamples

logProb

protected final boolean logProb

average

protected final boolean average

Constructor Detail

VAEReconProbScoreCalculator

public VAEReconProbScoreCalculator(org.nd4j.linalg.dataset.api.iterator.DataSetIterator iterator,
                                   int reconstructionProbNumSamples,
                                   boolean logProb)

Constructor for average reconstruction probability

Parameters:

iterator - Iterator

reconstructionProbNumSamples - Number of samples. See VariationalAutoencoder.reconstructionProbability(INDArray, int) for details

logProb - If true: calculate (negative) log probability. False: probability

VAEReconProbScoreCalculator

public VAEReconProbScoreCalculator(org.nd4j.linalg.dataset.api.iterator.DataSetIterator iterator,
                                   int reconstructionProbNumSamples,
                                   boolean logProb,
                                   boolean average)

Constructor for reconstruction probability

Parameters:

iterator - Iterator

reconstructionProbNumSamples - Number of samples. See VariationalAutoencoder.reconstructionProbability(INDArray, int) for details

logProb - If true: calculate (negative) log probability. False: probability

average - If true: return average (log) probability. False: sum of log probability.

Method Detail

reset

protected void reset()

Specified by:

reset in class BaseScoreCalculator<Model>

output

protected org.nd4j.linalg.api.ndarray.INDArray output(Model network,
                                                      org.nd4j.linalg.api.ndarray.INDArray input,
                                                      org.nd4j.linalg.api.ndarray.INDArray fMask,
                                                      org.nd4j.linalg.api.ndarray.INDArray lMask)

Specified by:

output in class BaseScoreCalculator<Model>

output

protected org.nd4j.linalg.api.ndarray.INDArray[] output(Model network,
                                                        org.nd4j.linalg.api.ndarray.INDArray[] input,
                                                        org.nd4j.linalg.api.ndarray.INDArray[] fMask,
                                                        org.nd4j.linalg.api.ndarray.INDArray[] lMask)

Specified by:

output in class BaseScoreCalculator<Model>

scoreMinibatch

protected double scoreMinibatch(Model net,
                                org.nd4j.linalg.api.ndarray.INDArray features,
                                org.nd4j.linalg.api.ndarray.INDArray labels,
                                org.nd4j.linalg.api.ndarray.INDArray fMask,
                                org.nd4j.linalg.api.ndarray.INDArray lMask,
                                org.nd4j.linalg.api.ndarray.INDArray output)

Overrides:

scoreMinibatch in class BaseScoreCalculator<Model>

scoreMinibatch

protected double scoreMinibatch(Model network,
                                org.nd4j.linalg.api.ndarray.INDArray[] features,
                                org.nd4j.linalg.api.ndarray.INDArray[] labels,
                                org.nd4j.linalg.api.ndarray.INDArray[] fMask,
                                org.nd4j.linalg.api.ndarray.INDArray[] lMask,
                                org.nd4j.linalg.api.ndarray.INDArray[] output)

Specified by:

scoreMinibatch in class BaseScoreCalculator<Model>

finalScore

protected double finalScore(double scoreSum,
                            int minibatchCount,
                            int exampleCount)

Specified by:

finalScore in class BaseScoreCalculator<Model>