hex.tree.gbm

Class GBMModel.GBMParameters

java.lang.Object

water.Iced<hex.Model.Parameters>

hex.Model.Parameters

hex.tree.SharedTreeModel.SharedTreeParameters

hex.tree.gbm.GBMModel.GBMParameters

All Implemented Interfaces:

hex.Model.AdaptFrameParameters, hex.Model.GetNTrees, CalibrationHelper.ParamsWithCalibration, java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, water.Freezable<hex.Model.Parameters>

Enclosing class:

GBMModel

public static class GBMModel.GBMParameters
extends SharedTreeModel.SharedTreeParameters

See Also:

Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from class hex.tree.SharedTreeModel.SharedTreeParameters

SharedTreeModel.SharedTreeParameters.HistogramType

Nested classes/interfaces inherited from class hex.Model.Parameters

hex.Model.Parameters.CategoricalEncodingScheme, hex.Model.Parameters.FoldAssignmentScheme

Field Summary

Fields

Modifier and Type	Field and Description
double	_col_sample_rate
java.lang.String[][]	_interaction_constraints
double	_learn_rate
double	_learn_rate_annealing
double	_max_abs_leafnode_pred
hex.KeyValue[]	_monotone_constraints
double	_pred_noise_bandwidth

Fields inherited from class hex.tree.SharedTreeModel.SharedTreeParameters

_build_tree_one_node, _calibrate_model, _calibration_frame, _calibration_method, _col_sample_rate_change_per_level, _col_sample_rate_per_tree, _histogram_type, _in_training_checkpoints_dir, _in_training_checkpoints_tree_interval, _initial_score_interval, _max_depth, _min_rows, _min_split_improvement, _nbins, _nbins_cats, _nbins_top_level, _ntrees, _parallel_main_model_building, _r2_stopping, _sample_rate, _sample_rate_per_class, _score_interval, _score_tree_interval, _use_best_cv_iteration

Fields inherited from class hex.Model.Parameters

_auc_type, _auto_rebalance, _auuc_nbins, _auuc_type, _balance_classes, _categorical_encoding, _check_constant_response, _checkpoint, _class_sampling_factors, _custom_distribution_func, _custom_metric_func, _cv_fold, _distribution, _export_checkpoints_dir, _fold_assignment, _fold_column, _gainslift_bins, _huber_alpha, _ignore_const_cols, _ignored_columns, _is_cv_model, _keep_cross_validation_fold_assignment, _keep_cross_validation_models, _keep_cross_validation_predictions, _keep_cross_validation_predictions_precision, _main_model_time_budget_factor, _max_after_balance_size, _max_categorical_levels, _max_confusion_matrix_size, _max_runtime_secs, _nfolds, _offset_column, _parallelize_cross_validation, _preprocessors, _pretrained_autoencoder, _quantile_alpha, _response_column, _score_each_iteration, _seed, _stopping_metric, _stopping_rounds, _stopping_tolerance, _train, _treatment_column, _tweedie_power, _valid, _weights_column, MAX_SUPPORTED_LEVELS

Constructor Summary

Constructors

Constructor and Description
GBMParameters()

Method Summary

Modifier and Type	Method and Description
java.lang.String	algoName()
Constraints	constraints(water.fvec.Frame f)
boolean	forceStrictlyReproducibleHistograms() Do we need to enable strictly deterministic way of building histograms? Used eg.
java.lang.String	fullName()
BranchInteractionConstraints	initialInteractionConstraints(GlobalInteractionConstraints ics)
GlobalInteractionConstraints	interactionConstraints(water.fvec.Frame frame)
java.lang.String	javaName()
boolean	useColSampling()

Methods inherited from class hex.tree.SharedTreeModel.SharedTreeParameters

calibrateModel, getCalibrationFrame, getCalibrationMethod, getNTrees, getParams, isStochastic, progressUnits, setCalibrationMethod, useRowSampling

Methods inherited from class hex.Model.Parameters

checksum, checksum, defaultDropConsCols, defaultStoppingTolerance, getCategoricalEncoding, getDependentKeys, getDistributionFamily, getFoldColumn, getMaxCategoricalLevels, getNonPredictors, getOffsetColumn, getOrMakeRealSeed, getResponseColumn, getTreatmentColumn, getUsedColumns, getWeightsColumn, hasCheckpoint, hasCustomMetricFunc, missingColumnsType, read_lock_frames, read_unlock_frames, setDistributionFamily, setTrain, train, valid

Methods inherited from class water.Iced

asBytes, clone, copyOver, frozenType, read, readExternal, readJSON, reloadFromBytes, toJsonBytes, toJsonString, write, writeExternal, writeJSON

Methods inherited from class java.lang.Object

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

Field Detail

_learn_rate

public double _learn_rate

_learn_rate_annealing

public double _learn_rate_annealing

_col_sample_rate

public double _col_sample_rate

_max_abs_leafnode_pred

public double _max_abs_leafnode_pred

_pred_noise_bandwidth

public double _pred_noise_bandwidth

_monotone_constraints

public hex.KeyValue[] _monotone_constraints

_interaction_constraints

public java.lang.String[][] _interaction_constraints

Constructor Detail

GBMParameters

public GBMParameters()

Method Detail

useColSampling

public boolean useColSampling()

Overrides:

useColSampling in class SharedTreeModel.SharedTreeParameters

algoName

public java.lang.String algoName()

Specified by:

algoName in class hex.Model.Parameters

fullName

public java.lang.String fullName()

Specified by:

fullName in class hex.Model.Parameters

javaName

public java.lang.String javaName()

Specified by:

javaName in class hex.Model.Parameters

forceStrictlyReproducibleHistograms

public boolean forceStrictlyReproducibleHistograms()

Description copied from class: SharedTreeModel.SharedTreeParameters

Do we need to enable strictly deterministic way of building histograms? Used eg. when monotonicity constraints in GBM are enabled, by default disabled

Overrides:

forceStrictlyReproducibleHistograms in class SharedTreeModel.SharedTreeParameters

Returns:

true if histograms should be built in deterministic way

constraints

public Constraints constraints(water.fvec.Frame f)

interactionConstraints

public GlobalInteractionConstraints interactionConstraints(water.fvec.Frame frame)

initialInteractionConstraints

public BranchInteractionConstraints initialInteractionConstraints(GlobalInteractionConstraints ics)