org.tribuo.classification.libsvm

Class LibSVMClassificationTrainer

java.lang.Object

org.tribuo.common.libsvm.LibSVMTrainer<Label>

org.tribuo.classification.libsvm.LibSVMClassificationTrainer

All Implemented Interfaces:

com.oracle.labs.mlrg.olcut.config.Configurable, com.oracle.labs.mlrg.olcut.provenance.Provenancable<TrainerProvenance>, WeightedLabels, Trainer<Label>

public class LibSVMClassificationTrainer
extends LibSVMTrainer<Label>
implements WeightedLabels

A trainer for classification models that uses LibSVM.

See:

 Chang CC, Lin CJ.
 "LIBSVM: a library for Support Vector Machines"
 ACM transactions on intelligent systems and technology (TIST), 2011.
 

for the nu-svc algorithm:

 Schölkopf B, Smola A, Williamson R, Bartlett P L.
 "New support vector algorithms"
 Neural Computation, 2000, 1207-1245.
 

and for the original algorithm:

 Cortes C, Vapnik V.
 "Support-Vector Networks"
 Machine Learning, 1995.
 

Field Summary

Fields inherited from class org.tribuo.common.libsvm.LibSVMTrainer

parameters, svmType

Fields inherited from interface org.tribuo.Trainer

DEFAULT_SEED

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	LibSVMClassificationTrainer()
	LibSVMClassificationTrainer(SVMParameters<Label> parameters)

Method Summary

Modifier and Type	Method and Description
protected LibSVMModel<Label>	createModel(ModelProvenance provenance, ImmutableFeatureMap featureIDMap, ImmutableOutputInfo<Label> outputIDInfo, List<libsvm.svm_model> models) Construct the appropriate subtype of LibSVMModel for the prediction task.
protected com.oracle.labs.mlrg.olcut.util.Pair<libsvm.svm_node[][],double[][]>	extractData(Dataset<Label> data, ImmutableOutputInfo<Label> outputInfo, ImmutableFeatureMap featureMap) Extracts the features and Outputs in LibLinear's format.
void	postConfig() Used by the OLCUT configuration system, and should not be called by external code.
void	setLabelWeights(Map<Label,Float> weights) Sets the label weights used by this trainer.
protected libsvm.svm_parameter	setupParameters(ImmutableOutputInfo<Label> outputIDInfo) Constructs the svm_parameter.
protected List<libsvm.svm_model>	trainModels(libsvm.svm_parameter curParams, int numFeatures, libsvm.svm_node[][] features, double[][] outputs) Train all the liblinear instances necessary for this dataset.

Methods inherited from class org.tribuo.common.libsvm.LibSVMTrainer

exampleToNodes, getInvocationCount, getProvenance, toString, train, train

Methods inherited from class java.lang.Object

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

Constructor Detail

LibSVMClassificationTrainer

protected LibSVMClassificationTrainer()

LibSVMClassificationTrainer

public LibSVMClassificationTrainer(SVMParameters<Label> parameters)

Method Detail

postConfig

public void postConfig()

Used by the OLCUT configuration system, and should not be called by external code.

Specified by:

postConfig in interface com.oracle.labs.mlrg.olcut.config.Configurable

Overrides:

postConfig in class LibSVMTrainer<Label>

createModel

protected LibSVMModel<Label> createModel(ModelProvenance provenance,
                                         ImmutableFeatureMap featureIDMap,
                                         ImmutableOutputInfo<Label> outputIDInfo,
                                         List<libsvm.svm_model> models)

Description copied from class: LibSVMTrainer

Construct the appropriate subtype of LibSVMModel for the prediction task.

Specified by:

createModel in class LibSVMTrainer<Label>

Parameters:

provenance - The model provenance.

featureIDMap - The feature id map.

outputIDInfo - The output id info.

models - The svm models.

Returns:

An implementation of LibSVMModel.

trainModels

protected List<libsvm.svm_model> trainModels(libsvm.svm_parameter curParams,
                                             int numFeatures,
                                             libsvm.svm_node[][] features,
                                             double[][] outputs)

Description copied from class: LibSVMTrainer

Train all the liblinear instances necessary for this dataset.

Specified by:

trainModels in class LibSVMTrainer<Label>

Parameters:

curParams - The LibLinear parameters.

numFeatures - The number of features in this dataset.

features - The features themselves.

outputs - The outputs.

Returns:

A list of liblinear models.

extractData

protected com.oracle.labs.mlrg.olcut.util.Pair<libsvm.svm_node[][],double[][]> extractData(Dataset<Label> data,
                                                                                           ImmutableOutputInfo<Label> outputInfo,
                                                                                           ImmutableFeatureMap featureMap)

Description copied from class: LibSVMTrainer

Extracts the features and Outputs in LibLinear's format.

Specified by:

extractData in class LibSVMTrainer<Label>

Parameters:

data - The input data.

outputInfo - The output info.

featureMap - The feature info.

Returns:

The features and outputs.

setupParameters

protected libsvm.svm_parameter setupParameters(ImmutableOutputInfo<Label> outputIDInfo)

Description copied from class: LibSVMTrainer

Constructs the svm_parameter. Most of the time this is a no-op, but classification overrides it to incorporate label weights if they exist.

Overrides:

setupParameters in class LibSVMTrainer<Label>

Parameters:

outputIDInfo - The output info.

Returns:

The svm_parameters to use for training.

setLabelWeights

public void setLabelWeights(Map<Label,Float> weights)

Description copied from interface: WeightedLabels

Sets the label weights used by this trainer.

Supply Collections.emptyMap() to turn off label weights.

Specified by:

setLabelWeights in interface WeightedLabels

Parameters:

weights - A map from Label instances to weight values.