weka.filters.unsupervised.attribute

Class Dl4jMlpFilter

java.lang.Object

weka.filters.Filter

weka.filters.SimpleFilter

weka.filters.SimpleBatchFilter

weka.filters.unsupervised.attribute.Dl4jMlpFilter

All Implemented Interfaces:

java.io.Serializable, weka.core.CapabilitiesHandler, weka.core.CapabilitiesIgnorer, weka.core.CommandlineRunnable, weka.core.OptionHandler, weka.core.RevisionHandler

public class Dl4jMlpFilter
extends weka.filters.SimpleBatchFilter
implements weka.core.OptionHandler, weka.core.CapabilitiesHandler

Weka filter that uses a neural network trained via Dl4jMlpClassifier as feature transformation.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
Dl4jMlpFilter() FILTER CODE.

Method Summary

Modifier and Type	Method and Description
void	addTransformationLayerName(java.lang.String transformationLayerName) Adds a new transformation layer for the filter to use.
boolean	allowAccessToFullInputFormat()
void	clearTransformationLayers() Clear the transformation layers to be used by the filter.
weka.core.Capabilities	getCapabilities() Returns default capabilities of the classifier.
AbstractInstanceIterator	getInstanceIterator()
java.lang.String[]	getOptions() Gets the current settings of the Classifier.
PoolingType	getPoolingType()
java.io.File	getSerializedModelFile() GET/SET METHODS
DenseLayer	getTransformationLayer(int index)
DenseLayer[]	getTransformationLayers()
boolean	getUseDefaultFeatureLayer()
AbstractZooModel	getZooModelType()
java.lang.String	globalInfo()
java.util.Enumeration<weka.core.Option>	listOptions() Returns an enumeration describing the available options.
void	setInstanceIterator(AbstractInstanceIterator instanceIterator)
void	setOptions(java.lang.String[] options) Parses a given list of options.
void	setPoolingType(PoolingType poolingType)
void	setSerializedModelFile(java.io.File modelPath)
void	setTransformationLayerNames(java.lang.String[] transformationLayerNames)
void	setTransformationLayers(DenseLayer[] transformationLayers)
void	setUseDefaultFeatureLayer(boolean useDefaultFeatureLayer)
void	setZooModelType(AbstractZooModel zooModelType) Clear the old transformation layers and set the new one if we've changed to a different model type.
java.lang.String[]	transformationLayersToNames() Return transformation layer names.

Methods inherited from class weka.filters.SimpleBatchFilter

batchFinished, input, input

Methods inherited from class weka.filters.SimpleFilter

setInputFormat

Methods inherited from class weka.filters.Filter

batchFilterFile, debugTipText, doNotCheckCapabilitiesTipText, filterFile, getCapabilities, getCopyOfInputFormat, getDebug, getDoNotCheckCapabilities, getOutputFormat, getRevision, isFirstBatchDone, isNewBatch, isOutputFormatDefined, main, makeCopies, makeCopy, mayRemoveInstanceAfterFirstBatchDone, numPendingOutput, output, outputPeek, postExecution, preExecution, run, runFilter, setDebug, setDoNotCheckCapabilities, toString, useFilter, wekaStaticWrapper

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, wait, wait, wait

Methods inherited from interface weka.core.OptionHandler

makeCopy

Constructor Detail

Dl4jMlpFilter

public Dl4jMlpFilter()

FILTER CODE.

Method Detail

getSerializedModelFile

@OptionMetadata(description="The trained Dl4jMlpClassifier object that contains the network, used for transformation.",
                displayName="Serialized model file",
                commandLineParamName="model",
                commandLineParamSynopsis="-model <File>",
                displayOrder=1)
public java.io.File getSerializedModelFile()

GET/SET METHODS

setSerializedModelFile

public void setSerializedModelFile(java.io.File modelPath)

getZooModelType

@OptionMetadata(description="The pretrained model from the DL4J Model Zoo (or a keras model)",
                displayName="Pretrained zoo model",
                commandLineParamName="zooModel",
                commandLineParamSynopsis="-zooModel <Model Zoo specification>",
                displayOrder=2)
public AbstractZooModel getZooModelType()

setZooModelType

public void setZooModelType(AbstractZooModel zooModelType)

Clear the old transformation layers and set the new one if we've changed to a different model type. This function is necessary from a usability perspective - if a user is switching between lots of different models, the final model's feature extraction layer should be applied. This requires clearing the old layers and applying the new one. The issue is if a user has selected a zoo model and wants to use some custom layers - when this method is called by the GUI, it will clear the 'old' layers (which include the custom applied layer) so the user can't actually use their custom setting. To fix this, a GUI flag is shown which the user can set to false if they wish to use non-default feature layers.

Parameters:

zooModelType - Type we're changing to

getInstanceIterator

@OptionMetadata(description="The instance iterator to use.",
                displayName="instance iterator",
                commandLineParamName="iterator",
                commandLineParamSynopsis="-iterator <string>")
public AbstractInstanceIterator getInstanceIterator()

setInstanceIterator

public void setInstanceIterator(AbstractInstanceIterator instanceIterator)

getPoolingType

@OptionMetadata(description="Pooling function to apply on intermediary activations",
                displayName="Pooling Type",
                commandLineParamName="poolingType",
                commandLineParamSynopsis="-poolingType <String>")
public PoolingType getPoolingType()

setPoolingType

public void setPoolingType(PoolingType poolingType)

getTransformationLayers

@OptionMetadata(description="Layers used for the feature transformation (can be left blank, default will be applied)",
                displayName="Feature extraction layers",
                commandLineParamName="layer-extract",
                commandLineParamSynopsis="-layer-extract <Dense Layer Spec>",
                displayOrder=0)
public DenseLayer[] getTransformationLayers()

getTransformationLayer

public DenseLayer getTransformationLayer(int index)

setTransformationLayers

public void setTransformationLayers(DenseLayer[] transformationLayers)

setTransformationLayerNames

public void setTransformationLayerNames(java.lang.String[] transformationLayerNames)

addTransformationLayerName

public void addTransformationLayerName(java.lang.String transformationLayerName)

Adds a new transformation layer for the filter to use.

Parameters:

transformationLayerName - name of the layer in the model to take activations from

clearTransformationLayers

public void clearTransformationLayers()

Clear the transformation layers to be used by the filter.

getUseDefaultFeatureLayer

public boolean getUseDefaultFeatureLayer()

setUseDefaultFeatureLayer

@OptionMetadata(displayName="Use default feature layer",
                description="Set to true to load the default feature extraction layer in the GUI - false to set your own custom layers.",
                displayOrder=0,
                commandLineParamName="default-feature-layer",
                commandLineParamSynopsis="-default-feature-layer",
                commandLineParamIsFlag=true)
public void setUseDefaultFeatureLayer(boolean useDefaultFeatureLayer)

allowAccessToFullInputFormat

public boolean allowAccessToFullInputFormat()

Overrides:

allowAccessToFullInputFormat in class weka.filters.SimpleBatchFilter

transformationLayersToNames

public java.lang.String[] transformationLayersToNames()

Return transformation layer names.

Returns:

String[] containing the names of transformation layers this filter is using

globalInfo

public java.lang.String globalInfo()

Specified by:

globalInfo in class weka.filters.SimpleFilter

getCapabilities

public weka.core.Capabilities getCapabilities()

Returns default capabilities of the classifier.

Specified by:

getCapabilities in interface weka.core.CapabilitiesHandler

Overrides:

getCapabilities in class weka.filters.Filter

Returns:

the capabilities of this classifier

listOptions

public java.util.Enumeration<weka.core.Option> listOptions()

Returns an enumeration describing the available options.

Specified by:

listOptions in interface weka.core.OptionHandler

Overrides:

listOptions in class weka.filters.Filter

Returns:

an enumeration of all the available options.

getOptions

public java.lang.String[] getOptions()

Gets the current settings of the Classifier.

Specified by:

getOptions in interface weka.core.OptionHandler

Overrides:

getOptions in class weka.filters.Filter

Returns:

an array of strings suitable for passing to setOptions

setOptions

public void setOptions(java.lang.String[] options)
                throws java.lang.Exception

Parses a given list of options.

Specified by:

setOptions in interface weka.core.OptionHandler

Overrides:

setOptions in class weka.filters.Filter

Parameters:

options - the list of options as an array of strings

Throws:

java.lang.Exception - if an option is not supported