Package ciir.umass.edu.learning.tree

Class MART

java.lang.Object

ciir.umass.edu.learning.Ranker

ciir.umass.edu.learning.tree.LambdaMART

ciir.umass.edu.learning.tree.MART

public class MART
extends LambdaMART

Author:

vdang This class implements MART for (point-wise) ranking: J.H. Friedman. Greedy function approximation: A gradient boosting machine. Technical Report, IMS Reitz Lecture, Stanford, 1999; see also Annals of Statistics, 2001.

Field Summary

Fields inherited from class ciir.umass.edu.learning.tree.LambdaMART

bestModelOnValidation, ensemble, gcCycle, hist, learningRate, martSamples, minLeafSupport, modelScores, modelScoresOnValidation, nRoundToStopEarly, nThreshold, nTreeLeaves, nTrees, pseudoResponses, sortedIdx, thresholds, weights

Fields inherited from class ciir.umass.edu.learning.Ranker

bestScoreOnValidationData, features, samples, scoreOnTrainingData, scorer, validationSamples, verbose

Constructor Summary

Constructors

Constructor	Description
MART()
MART(java.util.List<RankList> samples, int[] features, MetricScorer scorer)

Method Summary

Modifier and Type	Method	Description
protected void	computePseudoResponses()
Ranker	createNew()
java.lang.String	name()
protected void	updateTreeOutput(RegressionTree rt)

Methods inherited from class ciir.umass.edu.learning.tree.LambdaMART

computeModelScoreOnTraining, computeModelScoreOnTraining, computeModelScoreOnValidation, computeModelScoreOnValidation, computePseudoResponses, eval, getEnsemble, init, learn, loadFromString, model, printParameters, rank, sortSamplesByFeature, sortSamplesByFeature, toString

Methods inherited from class ciir.umass.edu.learning.Ranker

copy, getFeatures, getScoreOnTrainingData, getScoreOnValidationData, PRINT, PRINT, PRINT_MEMORY_USAGE, PRINTLN, PRINTLN, PRINTTIME, rank, rank, save, setFeatures, setMetricScorer, setTrainingSet, setValidationSet

Methods inherited from class java.lang.Object

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

Constructor Detail

MART

public MART()

MART

public MART(java.util.List<RankList> samples,
            int[] features,
            MetricScorer scorer)

Method Detail

createNew

public Ranker createNew()

Overrides:

createNew in class LambdaMART

name

public java.lang.String name()

Overrides:

name in class LambdaMART

computePseudoResponses

protected void computePseudoResponses()

Overrides:

computePseudoResponses in class LambdaMART

updateTreeOutput

protected void updateTreeOutput(RegressionTree rt)

Overrides:

updateTreeOutput in class LambdaMART