Package org.tribuo.classification.baseline

Class DummyClassifierTrainer

java.lang.Object

org.tribuo.classification.baseline.DummyClassifierTrainer

All Implemented Interfaces:

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

public final class DummyClassifierTrainer
extends Object
implements Trainer<Label>

A trainer for simple baseline classifiers. Use this only for comparison purposes, if you can't beat these baselines, your ML system doesn't work.

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static enum	DummyClassifierTrainer.DummyType	Types of dummy classifier.

Field Summary

Fields inherited from interface org.tribuo.Trainer

DEFAULT_SEED, INCREMENT_INVOCATION_COUNT

Method Summary

Modifier and Type	Method	Description
static DummyClassifierTrainer	createConstantTrainer(String constantLabel)	Creates a trainer which creates models which return a fixed label.
static DummyClassifierTrainer	createMostFrequentTrainer()	Creates a trainer which creates models which return a fixed label, the one which was most frequent in the training data.
static DummyClassifierTrainer	createStratifiedTrainer(long seed)	Creates a trainer which creates models which return random labels sampled from the training label distribution.
static DummyClassifierTrainer	createUniformTrainer(long seed)	Creates a trainer which creates models which return random labels sampled uniformly from the labels seen at training time.
int	getInvocationCount()
TrainerProvenance	getProvenance()
void	postConfig()	Used by the OLCUT configuration system, and should not be called by external code.
void	setInvocationCount(int invocationCount)
String	toString()
Model<Label>	train(Dataset<Label> examples, Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> instanceProvenance)
Model<Label>	train(Dataset<Label> examples, Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> instanceProvenance, int invocationCount)

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

Method Details

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

train

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

Specified by:

train in interface Trainer<Label>

train

public Model<Label> train(Dataset<Label> examples,
 Map<String,com.oracle.labs.mlrg.olcut.provenance.Provenance> instanceProvenance,
 int invocationCount)

Specified by:

train in interface Trainer<Label>

getInvocationCount

public int getInvocationCount()

Specified by:

getInvocationCount in interface Trainer<Label>

setInvocationCount

public void setInvocationCount(int invocationCount)

Specified by:

setInvocationCount in interface Trainer<Label>

toString

public String toString()

Overrides:

toString in class Object

getProvenance

public TrainerProvenance getProvenance()

Specified by:

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

createStratifiedTrainer

public static DummyClassifierTrainer createStratifiedTrainer(long seed)

Creates a trainer which creates models which return random labels sampled from the training label distribution.

Parameters:

seed - The RNG seed to use.

Returns:

A classification trainer.

createConstantTrainer

public static DummyClassifierTrainer createConstantTrainer(String constantLabel)

Creates a trainer which creates models which return a fixed label.

Parameters:

constantLabel - The label to return.

Returns:

A classification trainer.

createUniformTrainer

public static DummyClassifierTrainer createUniformTrainer(long seed)

Creates a trainer which creates models which return random labels sampled uniformly from the labels seen at training time.

Parameters:

seed - The RNG seed to use.

Returns:

A classification trainer.

createMostFrequentTrainer

public static DummyClassifierTrainer createMostFrequentTrainer()

Creates a trainer which creates models which return a fixed label, the one which was most frequent in the training data.

Returns:

A classification trainer.