org.apache.ignite.ml.tree.boosting

Class GDBOnTreesLearningStrategy

java.lang.Object

org.apache.ignite.ml.composition.boosting.GDBLearningStrategy

org.apache.ignite.ml.tree.boosting.GDBOnTreesLearningStrategy

public class GDBOnTreesLearningStrategy
extends GDBLearningStrategy

Gradient boosting on trees specific learning strategy reusing learning dataset with index between several learning iterations.

Field Summary

Fields inherited from class org.apache.ignite.ml.composition.boosting.GDBLearningStrategy

baseMdlTrainerBuilder, checkConvergenceStgyFactory, cntOfIterations, compositionWeights, envBuilder, externalLbToInternalMapping, loss, meanLbVal, sampleSize, trainerEnvironment

Constructor Summary

Constructors

Constructor and Description
GDBOnTreesLearningStrategy(boolean useIdx) Create an instance of learning strategy.

Method Summary

Modifier and Type	Method and Description
<K,V> List<IgniteModel<Vector,Double>>	update(GDBTrainer.GDBModel mdlToUpdate, DatasetBuilder<K,V> datasetBuilder, Preprocessor<K,V> vectorizer) Gets state of model in arguments, compare it with training parameters of trainer and if they are fit then trainer updates model in according to new data and return new model.

Methods inherited from class org.apache.ignite.ml.composition.boosting.GDBLearningStrategy

getCompositionWeights, getMeanValue, initLearningState, learnModels, withBaseModelTrainerBuilder, withCheckConvergenceStgyFactory, withCntOfIterations, withCompositionWeights, withDefaultGradStepSize, withEnvironmentBuilder, withExternalLabelToInternal, withLossGradient, withMeanLabelValue, withSampleSize

Methods inherited from class java.lang.Object

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

Constructor Detail

GDBOnTreesLearningStrategy

public GDBOnTreesLearningStrategy(boolean useIdx)

Create an instance of learning strategy.

Parameters:

useIdx - Use index.

Method Detail

update

public <K,V> List<IgniteModel<Vector,Double>> update(GDBTrainer.GDBModel mdlToUpdate,
                                                     DatasetBuilder<K,V> datasetBuilder,
                                                     Preprocessor<K,V> vectorizer)

Gets state of model in arguments, compare it with training parameters of trainer and if they are fit then trainer updates model in according to new data and return new model. In other case trains new model.

Overrides:

update in class GDBLearningStrategy

Type Parameters:

K - Type of a key in upstream data.

V - Type of a value in upstream data.

Parameters:

mdlToUpdate - Learned model.

datasetBuilder - Dataset builder.

vectorizer - Upstream preprocessor.

Returns:

Updated models list.