org.apache.ignite.ml.svm

Class SVMLinearMultiClassClassificationTrainer

java.lang.Object

org.apache.ignite.ml.svm.SVMLinearMultiClassClassificationTrainer

All Implemented Interfaces:

DatasetTrainer<SVMLinearMultiClassClassificationModel,Double>, SingleLabelDatasetTrainer<SVMLinearMultiClassClassificationModel>

public class SVMLinearMultiClassClassificationTrainer
extends Object
implements SingleLabelDatasetTrainer<SVMLinearMultiClassClassificationModel>

Base class for a soft-margin SVM linear multiclass-classification trainer based on the communication-efficient distributed dual coordinate ascent algorithm (CoCoA) with hinge-loss function. All common parameters are shared with bunch of binary classification trainers.

Constructor Summary

Constructors

Constructor and Description
SVMLinearMultiClassClassificationTrainer()

Method Summary

Modifier and Type	Method and Description
int	amountOfIterations() Gets the amount of outer iterations of SCDA algorithm.
int	amountOfLocIterations() Gets the amount of local iterations of SCDA algorithm.
<K,V> SVMLinearMultiClassClassificationModel	fit(DatasetBuilder<K,V> datasetBuilder, IgniteBiFunction<K,V,double[]> featureExtractor, IgniteBiFunction<K,V,Double> lbExtractor) Trains model based on the specified data.
double	lambda() Gets the regularization lambda.
SVMLinearMultiClassClassificationTrainer	withAmountOfIterations(int amountOfIterations) Set up the amount of outer iterations of SCDA algorithm.
SVMLinearMultiClassClassificationTrainer	withAmountOfLocIterations(int amountOfLocIterations) Set up the amount of local iterations of SCDA algorithm.
SVMLinearMultiClassClassificationTrainer	withLambda(double lambda) Set up the regularization parameter.

Methods inherited from class java.lang.Object

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

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

fit, fit

Constructor Detail

SVMLinearMultiClassClassificationTrainer

public SVMLinearMultiClassClassificationTrainer()

Method Detail

fit

public <K,V> SVMLinearMultiClassClassificationModel fit(DatasetBuilder<K,V> datasetBuilder,
                                                        IgniteBiFunction<K,V,double[]> featureExtractor,
                                                        IgniteBiFunction<K,V,Double> lbExtractor)

Trains model based on the specified data.

Specified by:

fit in interface DatasetTrainer<SVMLinearMultiClassClassificationModel,Double>

Type Parameters:

K - Type of a key in upstream data.

V - Type of a value in upstream data.

Parameters:

datasetBuilder - Dataset builder.

featureExtractor - Feature extractor.

lbExtractor - Label extractor.

Returns:

Model.

withLambda

public SVMLinearMultiClassClassificationTrainer withLambda(double lambda)

Set up the regularization parameter.

Parameters:

lambda - The regularization parameter. Should be more than 0.0.

Returns:

Trainer with new lambda parameter value.

lambda

public double lambda()

Gets the regularization lambda.

Returns:

The parameter value.

amountOfIterations

public int amountOfIterations()

Gets the amount of outer iterations of SCDA algorithm.

Returns:

The parameter value.

withAmountOfIterations

public SVMLinearMultiClassClassificationTrainer withAmountOfIterations(int amountOfIterations)

Set up the amount of outer iterations of SCDA algorithm.

Parameters:

amountOfIterations - The parameter value.

Returns:

Trainer with new amountOfIterations parameter value.

amountOfLocIterations

public int amountOfLocIterations()

Gets the amount of local iterations of SCDA algorithm.

Returns:

The parameter value.

withAmountOfLocIterations

public SVMLinearMultiClassClassificationTrainer withAmountOfLocIterations(int amountOfLocIterations)

Set up the amount of local iterations of SCDA algorithm.

Parameters:

amountOfLocIterations - The parameter value.

Returns:

Trainer with new amountOfLocIterations parameter value.