Package org.deeplearning4j.models.node2vec

Class Node2Vec.Builder<V extends SequenceElement,E extends Number>

java.lang.Object

org.deeplearning4j.models.sequencevectors.SequenceVectors.Builder<V>

org.deeplearning4j.models.node2vec.Node2Vec.Builder<V,E>

Enclosing class:

Node2Vec<V extends SequenceElement,E extends Number>

public static class Node2Vec.Builder<V extends SequenceElement,E extends Number>
extends SequenceVectors.Builder<V>

Field Summary

Fields inherited from class org.deeplearning4j.models.sequencevectors.SequenceVectors.Builder

batchSize, configuration, elementsLearningAlgorithm, enableScavenger, existingVectors, hugeModelExpected, intersectVectors, iterations, iterator, layerSize, learningRate, learningRateDecayWords, lockFactor, lookupTable, minLearningRate, minWordFrequency, modelUtils, negative, numEpochs, preciseMode, preciseWeightInit, resetModel, sampling, seed, sequenceLearningAlgorithm, STOP, stopWords, trainElementsVectors, trainSequenceVectors, UNK, unknownElement, useAdaGrad, useHierarchicSoftmax, useUnknown, variableWindows, vectorsListeners, vocabCache, vocabLimit, window, workers

Constructor Summary

Constructors

Constructor	Description
Builder(@NonNull GraphWalker<V> walker, @NonNull VectorsConfiguration configuration)

Method Summary

Modifier and Type	Method	Description
Node2Vec<V,E>	build()	Build SequenceVectors instance with defined settings/options
Node2Vec.Builder<V,E>	elementsLearningAlgorithm(@NonNull String algoName)	* Sets specific LearningAlgorithm as Elements Learning Algorithm
Node2Vec.Builder<V,E>	elementsLearningAlgorithm(@NonNull ElementsLearningAlgorithm<V> algorithm)	* Sets specific LearningAlgorithm as Elements Learning Algorithm
Node2Vec.Builder<V,E>	enableScavenger(boolean reallyEnable)	This method ebables/disables periodical vocab truncation during construction Default value: disabled
Node2Vec.Builder<V,E>	epochs(int numEpochs)	This method defines how much iterations should be done over whole training corpus during modelling
Node2Vec.Builder<V,E>	iterate(@NonNull SequenceIterator<V> iterator)	This method defines SequenceIterator to be used for model building
Node2Vec.Builder<V,E>	iterations(int iterations)	This method defines how much iterations should be done over batched sequences.
Node2Vec.Builder<V,E>	layerSize(int layerSize)	This method defines number of dimensions for outcome vectors.
Node2Vec.Builder<V,E>	learningRate(double learningRate)	This method defines initial learning rate.
Node2Vec.Builder<V,E>	lookupTable(@NonNull WeightLookupTable<V> lookupTable)	You can pass externally built WeightLookupTable, containing model weights and vocabulary.
Node2Vec.Builder<V,E>	minLearningRate(double minLearningRate)	This method defines minimum learning rate after decay being applied.
Node2Vec.Builder<V,E>	minWordFrequency(int minWordFrequency)	This method defines minimal element frequency for elements found in the training corpus.
Node2Vec.Builder<V,E>	modelUtils(@NonNull ModelUtils<V> modelUtils)	ModelUtils implementation, that will be used to access model.
Node2Vec.Builder<V,E>	negativeSample(double negative)	This method defines negative sampling value for skip-gram algorithm.
protected void	presetTables()	This method creates new WeightLookupTable and VocabCache if there were none set
Node2Vec.Builder<V,E>	resetModel(boolean reallyReset)	This method defines, should all model be reset before training.
Node2Vec.Builder<V,E>	sampling(double sampling)	This method defines sub-sampling threshold.
Node2Vec.Builder<V,E>	seed(long randomSeed)	Sets seed for random numbers generator.
Node2Vec.Builder<V,E>	sequenceLearningAlgorithm(@NonNull String algoName)	Sets specific LearningAlgorithm as Sequence Learning Algorithm
Node2Vec.Builder<V,E>	sequenceLearningAlgorithm(@NonNull SequenceLearningAlgorithm<V> algorithm)	Sets specific LearningAlgorithm as Sequence Learning Algorithm
Node2Vec.Builder<V,E>	setVectorsListeners(@NonNull Collection<VectorsListener<V>> vectorsListeners)	This method sets VectorsListeners for this SequenceVectors model
Node2Vec.Builder<V,E>	stopWords(@NonNull Collection<V> stopList)	You can provide collection of objects to be ignored, and excluded out of model Please note: Object labels and hashCode will be used for filtering
Node2Vec.Builder<V,E>	stopWords(@NonNull List<String> stopList)	You can provide collection of objects to be ignored, and excluded out of model Please note: Object labels and hashCode will be used for filtering
Node2Vec.Builder<V,E>	trainElementsRepresentation(boolean trainElements)
Node2Vec.Builder<V,E>	trainSequencesRepresentation(boolean trainSequences)
Node2Vec.Builder<V,E>	unknownElement(V element)	This method allows you to specify SequenceElement that will be used as UNK element, if UNK is used
Node2Vec.Builder<V,E>	useAdaGrad(boolean reallyUse)	This method defines if Adaptive Gradients should be used in calculations
protected Node2Vec.Builder<V,E>	useExistingWordVectors(@NonNull WordVectors vec)	This method allows you to use pre-built WordVectors model (e.g.
Node2Vec.Builder<V,E>	useHierarchicSoftmax(boolean reallyUse)	Enable/disable hierarchic softmax
Node2Vec.Builder<V,E>	usePreciseWeightInit(boolean reallyUse)	If set to true, initial weights for elements/sequences will be derived from elements themself.
Node2Vec.Builder<V,E>	useUnknown(boolean reallyUse)	This method allows you to specify, if UNK word should be used internally
Node2Vec.Builder<V,E>	useVariableWindow(int... windows)	This method allows to use variable window size.
Node2Vec.Builder<V,E>	vocabCache(@NonNull VocabCache<V> vocabCache)	You can pass externally built vocabCache object, containing vocabulary
Node2Vec.Builder<V,E>	windowSize(int windowSize)	Sets window size for skip-Gram training
Node2Vec.Builder<V,E>	workers(int numWorkers)	Sets number of worker threads to be used in calculations

Methods inherited from class org.deeplearning4j.models.sequencevectors.SequenceVectors.Builder

batchSize, intersectModel, limitVocabularySize, usePreciseMode

Methods inherited from class java.lang.Object

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

Constructor Detail

Builder

public Builder(@NonNull
               @NonNull GraphWalker<V> walker,
               @NonNull
               @NonNull VectorsConfiguration configuration)

Method Detail

useExistingWordVectors

protected Node2Vec.Builder<V,E> useExistingWordVectors(@NonNull
                                                             @NonNull WordVectors vec)

Description copied from class: SequenceVectors.Builder

This method allows you to use pre-built WordVectors model (e.g. SkipGram) for DBOW sequence learning. Existing model will be transferred into new model before training starts. PLEASE NOTE: This model has no effect for elements learning algorithms. Only sequence learning is affected. PLEASE NOTE: Non-normalized model is recommended to use here.

Overrides:

useExistingWordVectors in class SequenceVectors.Builder<V extends SequenceElement>

Parameters:

vec - existing WordVectors model

Returns:

iterate

public Node2Vec.Builder<V,E> iterate(@NonNull
                                           @NonNull SequenceIterator<V> iterator)

Description copied from class: SequenceVectors.Builder

This method defines SequenceIterator to be used for model building

Overrides:

iterate in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

sequenceLearningAlgorithm

public Node2Vec.Builder<V,E> sequenceLearningAlgorithm(@NonNull
                                                             @NonNull String algoName)

Description copied from class: SequenceVectors.Builder

Sets specific LearningAlgorithm as Sequence Learning Algorithm

Overrides:

sequenceLearningAlgorithm in class SequenceVectors.Builder<V extends SequenceElement>

Parameters:

algoName - fully qualified class name

Returns:

sequenceLearningAlgorithm

public Node2Vec.Builder<V,E> sequenceLearningAlgorithm(@NonNull
                                                             @NonNull SequenceLearningAlgorithm<V> algorithm)

Description copied from class: SequenceVectors.Builder

Sets specific LearningAlgorithm as Sequence Learning Algorithm

Overrides:

sequenceLearningAlgorithm in class SequenceVectors.Builder<V extends SequenceElement>

Parameters:

algorithm - SequenceLearningAlgorithm implementation

Returns:

elementsLearningAlgorithm

public Node2Vec.Builder<V,E> elementsLearningAlgorithm(@NonNull
                                                             @NonNull String algoName)

Description copied from class: SequenceVectors.Builder

* Sets specific LearningAlgorithm as Elements Learning Algorithm

Overrides:

elementsLearningAlgorithm in class SequenceVectors.Builder<V extends SequenceElement>

Parameters:

algoName - fully qualified class name

Returns:

elementsLearningAlgorithm

public Node2Vec.Builder<V,E> elementsLearningAlgorithm(@NonNull
                                                             @NonNull ElementsLearningAlgorithm<V> algorithm)

Description copied from class: SequenceVectors.Builder

* Sets specific LearningAlgorithm as Elements Learning Algorithm

Overrides:

elementsLearningAlgorithm in class SequenceVectors.Builder<V extends SequenceElement>

Parameters:

algorithm - ElementsLearningAlgorithm implementation

Returns:

iterations

public Node2Vec.Builder<V,E> iterations(int iterations)

Description copied from class: SequenceVectors.Builder

This method defines how much iterations should be done over batched sequences.

Overrides:

iterations in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

epochs

public Node2Vec.Builder<V,E> epochs(int numEpochs)

Description copied from class: SequenceVectors.Builder

This method defines how much iterations should be done over whole training corpus during modelling

Overrides:

epochs in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

workers

public Node2Vec.Builder<V,E> workers(int numWorkers)

Description copied from class: SequenceVectors.Builder

Sets number of worker threads to be used in calculations

Overrides:

workers in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

useHierarchicSoftmax

public Node2Vec.Builder<V,E> useHierarchicSoftmax(boolean reallyUse)

Description copied from class: SequenceVectors.Builder

Enable/disable hierarchic softmax

Overrides:

useHierarchicSoftmax in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

useAdaGrad

public Node2Vec.Builder<V,E> useAdaGrad(boolean reallyUse)

Description copied from class: SequenceVectors.Builder

This method defines if Adaptive Gradients should be used in calculations

Overrides:

useAdaGrad in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

layerSize

public Node2Vec.Builder<V,E> layerSize(int layerSize)

Description copied from class: SequenceVectors.Builder

This method defines number of dimensions for outcome vectors. Please note: This option has effect only if lookupTable wasn't defined during building process.

Overrides:

layerSize in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

learningRate

public Node2Vec.Builder<V,E> learningRate(double learningRate)

Description copied from class: SequenceVectors.Builder

This method defines initial learning rate. Default value is 0.025

Overrides:

learningRate in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

minWordFrequency

public Node2Vec.Builder<V,E> minWordFrequency(int minWordFrequency)

Description copied from class: SequenceVectors.Builder

This method defines minimal element frequency for elements found in the training corpus. All elements with frequency below this threshold will be removed before training. Please note: this method has effect only if vocabulary is built internally.

Overrides:

minWordFrequency in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

minLearningRate

public Node2Vec.Builder<V,E> minLearningRate(double minLearningRate)

Description copied from class: SequenceVectors.Builder

This method defines minimum learning rate after decay being applied. Default value is 0.01

Overrides:

minLearningRate in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

resetModel

public Node2Vec.Builder<V,E> resetModel(boolean reallyReset)

Description copied from class: SequenceVectors.Builder

This method defines, should all model be reset before training. If set to true, vocabulary and WeightLookupTable will be reset before training, and will be built from scratches

Overrides:

resetModel in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

vocabCache

public Node2Vec.Builder<V,E> vocabCache(@NonNull
                                              @NonNull VocabCache<V> vocabCache)

Description copied from class: SequenceVectors.Builder

You can pass externally built vocabCache object, containing vocabulary

Overrides:

vocabCache in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

lookupTable

public Node2Vec.Builder<V,E> lookupTable(@NonNull
                                               @NonNull WeightLookupTable<V> lookupTable)

Description copied from class: SequenceVectors.Builder

You can pass externally built WeightLookupTable, containing model weights and vocabulary.

Overrides:

lookupTable in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

sampling

public Node2Vec.Builder<V,E> sampling(double sampling)

Description copied from class: SequenceVectors.Builder

This method defines sub-sampling threshold.

Overrides:

sampling in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

negativeSample

public Node2Vec.Builder<V,E> negativeSample(double negative)

Description copied from class: SequenceVectors.Builder

This method defines negative sampling value for skip-gram algorithm.

Overrides:

negativeSample in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

stopWords

public Node2Vec.Builder<V,E> stopWords(@NonNull
                                             @NonNull List<String> stopList)

Description copied from class: SequenceVectors.Builder

You can provide collection of objects to be ignored, and excluded out of model Please note: Object labels and hashCode will be used for filtering

Overrides:

stopWords in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

trainElementsRepresentation

public Node2Vec.Builder<V,E> trainElementsRepresentation(boolean trainElements)

Overrides:

trainElementsRepresentation in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

trainSequencesRepresentation

public Node2Vec.Builder<V,E> trainSequencesRepresentation(boolean trainSequences)

Overrides:

trainSequencesRepresentation in class SequenceVectors.Builder<V extends SequenceElement>

stopWords

public Node2Vec.Builder<V,E> stopWords(@NonNull
                                             @NonNull Collection<V> stopList)

Description copied from class: SequenceVectors.Builder

You can provide collection of objects to be ignored, and excluded out of model Please note: Object labels and hashCode will be used for filtering

Overrides:

stopWords in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

windowSize

public Node2Vec.Builder<V,E> windowSize(int windowSize)

Description copied from class: SequenceVectors.Builder

Sets window size for skip-Gram training

Overrides:

windowSize in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

seed

public Node2Vec.Builder<V,E> seed(long randomSeed)

Description copied from class: SequenceVectors.Builder

Sets seed for random numbers generator. Please note: this has effect only if vocabulary and WeightLookupTable is built internally

Overrides:

seed in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

modelUtils

public Node2Vec.Builder<V,E> modelUtils(@NonNull
                                              @NonNull ModelUtils<V> modelUtils)

Description copied from class: SequenceVectors.Builder

ModelUtils implementation, that will be used to access model. Methods like: similarity, wordsNearest, accuracy are provided by user-defined ModelUtils

Overrides:

modelUtils in class SequenceVectors.Builder<V extends SequenceElement>

Parameters:

modelUtils - model utils to be used

Returns:

useUnknown

public Node2Vec.Builder<V,E> useUnknown(boolean reallyUse)

Description copied from class: SequenceVectors.Builder

This method allows you to specify, if UNK word should be used internally

Overrides:

useUnknown in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

unknownElement

public Node2Vec.Builder<V,E> unknownElement(@NonNull
                                                  V element)

Description copied from class: SequenceVectors.Builder

This method allows you to specify SequenceElement that will be used as UNK element, if UNK is used

Overrides:

unknownElement in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

useVariableWindow

public Node2Vec.Builder<V,E> useVariableWindow(int... windows)

Description copied from class: SequenceVectors.Builder

This method allows to use variable window size. In this case, every batch gets processed using one of predefined window sizes

Overrides:

useVariableWindow in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

usePreciseWeightInit

public Node2Vec.Builder<V,E> usePreciseWeightInit(boolean reallyUse)

Description copied from class: SequenceVectors.Builder

If set to true, initial weights for elements/sequences will be derived from elements themself. However, this implies additional cycle through input iterator. Default value: FALSE

Overrides:

usePreciseWeightInit in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

presetTables

protected void presetTables()

Description copied from class: SequenceVectors.Builder

This method creates new WeightLookupTable and VocabCache if there were none set

Overrides:

presetTables in class SequenceVectors.Builder<V extends SequenceElement>

setVectorsListeners

public Node2Vec.Builder<V,E> setVectorsListeners(@NonNull
                                                       @NonNull Collection<VectorsListener<V>> vectorsListeners)

Description copied from class: SequenceVectors.Builder

This method sets VectorsListeners for this SequenceVectors model

Overrides:

setVectorsListeners in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

enableScavenger

public Node2Vec.Builder<V,E> enableScavenger(boolean reallyEnable)

Description copied from class: SequenceVectors.Builder

This method ebables/disables periodical vocab truncation during construction Default value: disabled

Overrides:

enableScavenger in class SequenceVectors.Builder<V extends SequenceElement>

Returns:

build

public Node2Vec<V,E> build()

Description copied from class: SequenceVectors.Builder

Build SequenceVectors instance with defined settings/options

Overrides:

build in class SequenceVectors.Builder<V extends SequenceElement>

Returns: