org.tribuo.ensemble

Class BaggingTrainer<T extends Output<T>>

java.lang.Object

org.tribuo.ensemble.BaggingTrainer<T>

All Implemented Interfaces:

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

Direct Known Subclasses:

RandomForestTrainer

public class BaggingTrainer<T extends Output<T>>
extends Object
implements Trainer<T>

A Trainer that wraps another trainer and produces a bagged ensemble.

A bagged ensemble is a set of models each of which was trained on a bootstrap sample of the original dataset, combined with an unweighted majority vote.

See:

 J. Friedman, T. Hastie, & R. Tibshirani.
 "The Elements of Statistical Learning"
 Springer 2001. PDF
 

Field Summary

Fields

Modifier and Type	Field and Description
protected EnsembleCombiner<T>	combiner
protected Trainer<T>	innerTrainer
protected int	numMembers
protected SplittableRandom	rng
protected long	seed
protected int	trainInvocationCounter

Fields inherited from interface org.tribuo.Trainer

DEFAULT_SEED

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	BaggingTrainer() For the configuration system.
	BaggingTrainer(Trainer<T> trainer, EnsembleCombiner<T> combiner, int numMembers)
	BaggingTrainer(Trainer<T> trainer, EnsembleCombiner<T> combiner, int numMembers, long seed)

Method Summary

Modifier and Type	Method and Description
protected String	ensembleName()
int	getInvocationCount() The number of times this trainer instance has had it's train method invoked.
TrainerProvenance	getProvenance()
void	postConfig()
String	toString()
Model<T>	train(Dataset<T> examples, Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> runProvenance) Trains a predictive model using the examples in the given data set.
protected Model<T>	trainSingleModel(Dataset<T> examples, ImmutableFeatureMap featureIDs, ImmutableOutputInfo<T> labelIDs, SplittableRandom localRNG, Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> runProvenance)

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.tribuo.Trainer

train

Field Detail

innerTrainer

@Config(mandatory=true,
        description="The trainer to use for each ensemble member.")
protected Trainer<T extends Output<T>> innerTrainer

numMembers

@Config(mandatory=true,
        description="The number of ensemble members to train.")
protected int numMembers

seed

@Config(mandatory=true,
        description="The seed for the RNG.")
protected long seed

combiner

@Config(mandatory=true,
        description="The combination function to aggregate each ensemble member\'s outputs.")
protected EnsembleCombiner<T extends Output<T>> combiner

rng

protected SplittableRandom rng

trainInvocationCounter

protected int trainInvocationCounter

Constructor Detail

BaggingTrainer

protected BaggingTrainer()

For the configuration system.

BaggingTrainer

public BaggingTrainer(Trainer<T> trainer,
                      EnsembleCombiner<T> combiner,
                      int numMembers)

BaggingTrainer

public BaggingTrainer(Trainer<T> trainer,
                      EnsembleCombiner<T> combiner,
                      int numMembers,
                      long seed)

Method Detail

postConfig

public void postConfig()

Specified by:

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

ensembleName

protected String ensembleName()

toString

public String toString()

Overrides:

toString in class Object

train

public Model<T> train(Dataset<T> examples,
                      Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> runProvenance)

Description copied from interface: Trainer

Trains a predictive model using the examples in the given data set.

Specified by:

train in interface Trainer<T extends Output<T>>

Parameters:

examples - the data set containing the examples.

runProvenance - Training run specific provenance (e.g., fold number).

Returns:

a predictive model that can be used to generate predictions for new examples.

trainSingleModel

protected Model<T> trainSingleModel(Dataset<T> examples,
                                    ImmutableFeatureMap featureIDs,
                                    ImmutableOutputInfo<T> labelIDs,
                                    SplittableRandom localRNG,
                                    Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> runProvenance)

getInvocationCount

public int getInvocationCount()

Description copied from interface: Trainer

The number of times this trainer instance has had it's train method invoked.

This is used to determine how many times the trainer's RNG has been accessed to ensure replicability in the random number stream.

Specified by:

getInvocationCount in interface Trainer<T extends Output<T>>

Returns:

The number of train invocations.

getProvenance

public TrainerProvenance getProvenance()

Specified by:

getProvenance in interface com.oracle.labs.mlrg.olcut.provenance.Provenancable<TrainerProvenance>