mulan.classifier.transformation
Class CalibratedLabelRanking

java.lang.Object
  mulan.classifier.MultiLabelLearnerBase
      mulan.classifier.transformation.TransformationBasedMultiLabelLearner
          mulan.classifier.transformation.CalibratedLabelRanking

All Implemented Interfaces:

Serializable, MultiLabelLearner, TechnicalInformationHandler

public class CalibratedLabelRanking
extends TransformationBasedMultiLabelLearner

Class implementing the Calibrated Label Ranking (CLR) algorithm. For more information, see

Fuernkranz, Johannes, Huellermeier, Eyke, Loza Mencia, Eneldo, Brinker, Klaus (2008). Multilabel classification via calibrated label ranking. Machine Learning. 73(2):133--153.

BibTeX:

 @article{Fuernkranz2008,
    author = {Fuernkranz, Johannes and Huellermeier, Eyke and Loza Mencia, Eneldo and Brinker, Klaus},
    journal = {Machine Learning},
    number = {2},
    pages = {133--153},
    title = {Multilabel classification via calibrated label ranking},
    volume = {73},
    year = {2008}
 }
 

Version:

2012.02.27

Author:

Elise Rairat, Sang-Hyeun Park, Grigorios Tsoumakas

See Also:

Serialized Form

Field Summary

protected Instances[]	metaDataTest headers of the training sets of the one vs one models
protected boolean[]	nodata whether no data exist for one-vs-one learning
protected int	numModels number of one vs one models
protected Classifier[]	oneVsOneModels array holding the one vs one models
protected Instances	trainingdata temporary training data for each one vs one model
protected BinaryRelevance	virtualLabelModels binary relevance models for the virtual label

Fields inherited from class mulan.classifier.transformation.TransformationBasedMultiLabelLearner

baseClassifier

Fields inherited from class mulan.classifier.MultiLabelLearnerBase

featureIndices, labelIndices, numLabels

Constructor Summary

CalibratedLabelRanking()
Default constructor using J48 as underlying classifier

CalibratedLabelRanking(Classifier classifier)
Constructor that initializes the learner with a base algorithm

Method Summary

protected void	buildInternal(MultiLabelInstances trainingSet) Learner specific implementation of building the model from MultiLabelInstances training data set.
boolean	getStandardVoting() Get whether standard voting is turned on.
TechnicalInformation	getTechnicalInformation() Returns an instance of a TechnicalInformation object, containing detailed information about the technical background of this class, e.g., paper reference or book this class is based on.
String	globalInfo() Returns a string describing the classifier.
protected MultiLabelOutput	makePredictionInternal(Instance instance) This method does a prediction for an instance with the values of label missing Temporary included to switch between standard voting and qweighted multilabel voting
MultiLabelOutput	makePredictionQW(Instance instance) This method does a prediction for an instance with the values of label missing according to QWeighted algorithm for Multilabel Classification (QCMLPP2), which is described in : Loza Mencia, E., Park, S.
MultiLabelOutput	makePredictionStandard(Instance instance) This method does a prediction for an instance with the values of label missing
void	setStandardVoting(boolean standardVoting) Set Prediction to standard voting mode.

Methods inherited from class mulan.classifier.transformation.TransformationBasedMultiLabelLearner

getBaseClassifier

Methods inherited from class mulan.classifier.MultiLabelLearnerBase

build, debug, getDebug, isModelInitialized, isUpdatable, makeCopy, makePrediction, setDebug

Methods inherited from class java.lang.Object

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

Field Detail

oneVsOneModels

protected Classifier[] oneVsOneModels

array holding the one vs one models

numModels

protected int numModels

number of one vs one models

trainingdata

protected Instances trainingdata

temporary training data for each one vs one model

metaDataTest

protected Instances[] metaDataTest

headers of the training sets of the one vs one models

virtualLabelModels

protected BinaryRelevance virtualLabelModels

binary relevance models for the virtual label

nodata

protected boolean[] nodata

whether no data exist for one-vs-one learning

Constructor Detail

CalibratedLabelRanking

public CalibratedLabelRanking()

Default constructor using J48 as underlying classifier

CalibratedLabelRanking

public CalibratedLabelRanking(Classifier classifier)

Constructor that initializes the learner with a base algorithm

Parameters:

classifier - the binary classification algorithm to use

Method Detail

setStandardVoting

public void setStandardVoting(boolean standardVoting)

Set Prediction to standard voting mode.

Parameters:

standardVoting - true if standard voting should be used

getStandardVoting

public boolean getStandardVoting()

Get whether standard voting is turned on.

Returns:

true if standard voting is on

buildInternal

protected void buildInternal(MultiLabelInstances trainingSet)
                      throws Exception

Description copied from class: MultiLabelLearnerBase

Learner specific implementation of building the model from MultiLabelInstances training data set. This method is called from MultiLabelLearnerBase.build(MultiLabelInstances) method, where behavior common across all learners is applied.

Specified by:

buildInternal in class MultiLabelLearnerBase

Parameters:

trainingSet - the training data set

Throws:

Exception - if learner model was not created successfully

makePredictionInternal

protected MultiLabelOutput makePredictionInternal(Instance instance)
                                           throws Exception

This method does a prediction for an instance with the values of label missing Temporary included to switch between standard voting and qweighted multilabel voting

Specified by:

makePredictionInternal in class MultiLabelLearnerBase

Parameters:

instance -

Returns:

prediction

Throws:

Exception

InvalidDataException - if specified instance data is invalid and can not be processed by the learner

makePredictionStandard

public MultiLabelOutput makePredictionStandard(Instance instance)
                                        throws Exception

This method does a prediction for an instance with the values of label missing

Parameters:

instance -

Returns:

prediction

Throws:

Exception

makePredictionQW

public MultiLabelOutput makePredictionQW(Instance instance)
                                  throws Exception

This method does a prediction for an instance with the values of label missing according to QWeighted algorithm for Multilabel Classification (QCMLPP2), which is described in : Loza Mencia, E., Park, S.-H., and Fuernkranz, J. (2009) Efficient voting prediction for pairwise multilabel classification. In Proceedings of 17th European Symposium on Artificial Neural Networks (ESANN 2009), Bruges (Belgium), April 2009 This method reduces the number of classifier evaluations and guarantees the same Multilabel Output as ordinary Voting. But: the estimated confidences are only approximated. Therefore, ranking-based performances are worse than ordinary voting.

Parameters:

instance -

Returns:

prediction

Throws:

Exception

globalInfo

public String globalInfo()

Returns a string describing the classifier.

Overrides:

globalInfo in class TransformationBasedMultiLabelLearner

Returns:

a string description of the classifier

getTechnicalInformation

public TechnicalInformation getTechnicalInformation()

Description copied from class: TransformationBasedMultiLabelLearner

Returns an instance of a TechnicalInformation object, containing detailed information about the technical background of this class, e.g., paper reference or book this class is based on.

Specified by:

getTechnicalInformation in interface TechnicalInformationHandler

Overrides:

getTechnicalInformation in class TransformationBasedMultiLabelLearner

Returns:

the technical information about this class