org.apache.ignite.ml.composition.stacking

Class StackedVectorDatasetTrainer<O,AM extends IgniteModel<Vector,O>,L>

java.lang.Object

org.apache.ignite.ml.trainers.DatasetTrainer<StackedModel<IS,IA,O,AM>,L>

org.apache.ignite.ml.composition.stacking.StackedDatasetTrainer<I,I,O,AM,L>

org.apache.ignite.ml.composition.stacking.SimpleStackedDatasetTrainer<Vector,O,AM,L>

org.apache.ignite.ml.composition.stacking.StackedVectorDatasetTrainer<O,AM,L>

Type Parameters:

O - Type of aggregator model output.

L - Type of labels.

AM - Type of aggregator model.

public class StackedVectorDatasetTrainer<O,AM extends IgniteModel<Vector,O>,L>
extends SimpleStackedDatasetTrainer<Vector,O,AM,L>

StackedDatasetTrainer with Vector as submodels input and output.

Nested Class Summary

Nested classes/interfaces inherited from class org.apache.ignite.ml.trainers.DatasetTrainer

DatasetTrainer.EmptyDatasetException

Field Summary

Fields inherited from class org.apache.ignite.ml.trainers.DatasetTrainer

envBuilder, environment

Constructor Summary

Constructors

Constructor and Description
StackedVectorDatasetTrainer() Constructs instance of this class.
StackedVectorDatasetTrainer(DatasetTrainer<AM,L> aggregatingTrainer) Constructs instance of this class.

Method Summary

Modifier and Type	Method and Description
<M1 extends IgniteModel<Matrix,Matrix>> StackedVectorDatasetTrainer<O,AM,L>	addMatrix2MatrixTrainer(DatasetTrainer<M1,L> trainer) Shortcut for adding trainer Matrix -> Matrix where this trainer is treated as Vector -> Vector, where input Vector is turned into 1 x cols Matrix and output is a first row of output Matrix.
<M1 extends IgniteModel<Vector,Vector>> StackedVectorDatasetTrainer<O,AM,L>	addTrainer(DatasetTrainer<M1,L> trainer) Adds submodel trainer along with converters needed on training and inference stages.
<M1 extends IgniteModel<Vector,Double>> StackedVectorDatasetTrainer<O,AM,L>	addTrainerWithDoubleOutput(DatasetTrainer<M1,L> trainer) Shortcut for adding trainer Vector -> Double where this trainer is treated as Vector -> Vector, where output Vector is constructed by wrapping double value.
StackedVectorDatasetTrainer<O,AM,L>	withAggregatorInputMerger(IgniteBinaryOperator<Vector> merger) Specify binary operator used to merge submodels outputs to one.
StackedVectorDatasetTrainer<O,AM,L>	withAggregatorTrainer(DatasetTrainer<AM,L> aggregatorTrainer) Specify aggregator trainer.
<L1> StackedVectorDatasetTrainer<O,AM,L1>	withConvertedLabels(IgniteFunction<L1,L> new2Old) Creates DatasetTrainer with same training logic, but able to accept labels of given new type of labels.
StackedVectorDatasetTrainer<O,AM,L>	withEnvironmentBuilder(LearningEnvironmentBuilder envBuilder) Changes learning Environment.
StackedVectorDatasetTrainer<O,AM,L>	withOriginalFeaturesDropped() Drop original features during training and inference.
StackedVectorDatasetTrainer<O,AM,L>	withOriginalFeaturesKept() Keep original features using IgniteFunction.identity() as submodelInput2AggregatingInputConverter.
StackedVectorDatasetTrainer<O,AM,L>	withOriginalFeaturesKept(IgniteFunction<Vector,Vector> submodelInput2AggregatingInputConverter) Keep original features during training and propagate submodels input to aggregator during inference using given function.
StackedVectorDatasetTrainer<O,AM,L>	withSubmodelOutput2VectorConverter(IgniteFunction<Vector,Vector> submodelOutput2VectorConverter) Set function used for conversion of submodel output to Vector.
StackedVectorDatasetTrainer<O,AM,L>	withVector2SubmodelInputConverter(IgniteFunction<Vector,Vector> vector2SubmodelInputConverter) Set function used for conversion of Vector to submodel input.

Methods inherited from class org.apache.ignite.ml.composition.stacking.StackedDatasetTrainer

fitWithInitializedDeployingContext, isUpdateable, update, updateModel

Methods inherited from class org.apache.ignite.ml.trainers.DatasetTrainer

fit, fit, fit, fit, fit, fit, getLastTrainedModelOrThrowEmptyDatasetException, identityTrainer, learningEnvironment, update, update, update, update

Methods inherited from class java.lang.Object

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

Constructor Detail

StackedVectorDatasetTrainer

public StackedVectorDatasetTrainer(DatasetTrainer<AM,L> aggregatingTrainer)

Constructs instance of this class.

Parameters:

aggregatingTrainer - Aggregator trainer.

StackedVectorDatasetTrainer

public StackedVectorDatasetTrainer()

Constructs instance of this class.

Method Detail

addTrainer

public <M1 extends IgniteModel<Vector,Vector>> StackedVectorDatasetTrainer<O,AM,L> addTrainer(DatasetTrainer<M1,L> trainer)

Adds submodel trainer along with converters needed on training and inference stages.

Overrides:

addTrainer in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

trainer - Submodel trainer.

Returns:

This object.

withAggregatorTrainer

public StackedVectorDatasetTrainer<O,AM,L> withAggregatorTrainer(DatasetTrainer<AM,L> aggregatorTrainer)

Specify aggregator trainer.

Overrides:

withAggregatorTrainer in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

aggregatorTrainer - Aggregator trainer.

Returns:

This object.

withOriginalFeaturesKept

public StackedVectorDatasetTrainer<O,AM,L> withOriginalFeaturesKept()

Keep original features using IgniteFunction.identity() as submodelInput2AggregatingInputConverter.

Overrides:

withOriginalFeaturesKept in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Returns:

This object.

withOriginalFeaturesDropped

public StackedVectorDatasetTrainer<O,AM,L> withOriginalFeaturesDropped()

Drop original features during training and inference.

Overrides:

withOriginalFeaturesDropped in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Returns:

This object.

withOriginalFeaturesKept

public StackedVectorDatasetTrainer<O,AM,L> withOriginalFeaturesKept(IgniteFunction<Vector,Vector> submodelInput2AggregatingInputConverter)

Keep original features during training and propagate submodels input to aggregator during inference using given function. Note that if this object is on, training will be done on vector obtaining from concatenating features passed to submodels trainers and outputs of submodels converted to vectors, this can, for example influence aggregator model input vector dimension (if IS = Vector), or, more generally, some IS parameters which are not reflected just by its type. So converter should be written accordingly.

Overrides:

withOriginalFeaturesKept in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

submodelInput2AggregatingInputConverter - Function used to propagate submodels input to aggregator.

Returns:

This object.

withSubmodelOutput2VectorConverter

public StackedVectorDatasetTrainer<O,AM,L> withSubmodelOutput2VectorConverter(IgniteFunction<Vector,Vector> submodelOutput2VectorConverter)

Set function used for conversion of submodel output to Vector. This function is used during building of dataset for training aggregator model. This dataset is augmented with results of submodels converted to Vector.

Overrides:

withSubmodelOutput2VectorConverter in class StackedDatasetTrainer<Vector,Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

submodelOutput2VectorConverter - Function used for conversion of submodel output to Vector.

Returns:

This object.

withVector2SubmodelInputConverter

public StackedVectorDatasetTrainer<O,AM,L> withVector2SubmodelInputConverter(IgniteFunction<Vector,Vector> vector2SubmodelInputConverter)

Set function used for conversion of Vector to submodel input. This function is used during building of dataset for training aggregator model. This dataset is augmented with results of submodels applied to Vectors in original dataset.

Overrides:

withVector2SubmodelInputConverter in class StackedDatasetTrainer<Vector,Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

vector2SubmodelInputConverter - Function used for conversion of Vector to submodel input.

Returns:

This object.

withAggregatorInputMerger

public StackedVectorDatasetTrainer<O,AM,L> withAggregatorInputMerger(IgniteBinaryOperator<Vector> merger)

Specify binary operator used to merge submodels outputs to one.

Overrides:

withAggregatorInputMerger in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

merger - Binary operator used to merge submodels outputs to one.

Returns:

This object.

withEnvironmentBuilder

public StackedVectorDatasetTrainer<O,AM,L> withEnvironmentBuilder(LearningEnvironmentBuilder envBuilder)

Changes learning Environment.

Overrides:

withEnvironmentBuilder in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Parameters:

envBuilder - Learning environment builder.

withConvertedLabels

public <L1> StackedVectorDatasetTrainer<O,AM,L1> withConvertedLabels(IgniteFunction<L1,L> new2Old)

Creates DatasetTrainer with same training logic, but able to accept labels of given new type of labels.

Overrides:

withConvertedLabels in class SimpleStackedDatasetTrainer<Vector,O,AM extends IgniteModel<Vector,O>,L>

Type Parameters:

L1 - New labels type.

Parameters:

new2Old - Converter of new labels to old labels.

Returns:

DatasetTrainer with same training logic, but able to accept labels of given new type of labels.

addTrainerWithDoubleOutput

public <M1 extends IgniteModel<Vector,Double>> StackedVectorDatasetTrainer<O,AM,L> addTrainerWithDoubleOutput(DatasetTrainer<M1,L> trainer)

Shortcut for adding trainer Vector -> Double where this trainer is treated as Vector -> Vector, where output Vector is constructed by wrapping double value.

Type Parameters:

M1 - Type of submodel trainer model.

Parameters:

trainer - Submodel trainer.

Returns:

This object.

addMatrix2MatrixTrainer

public <M1 extends IgniteModel<Matrix,Matrix>> StackedVectorDatasetTrainer<O,AM,L> addMatrix2MatrixTrainer(DatasetTrainer<M1,L> trainer)

Shortcut for adding trainer Matrix -> Matrix where this trainer is treated as Vector -> Vector, where input Vector is turned into 1 x cols Matrix and output is a first row of output Matrix.

Type Parameters:

M1 - Type of submodel trainer model.

Parameters:

trainer - Submodel trainer.

Returns:

This object.