water.bindings.pojos

Class GLRMParametersV3

java.lang.Object

water.bindings.pojos.SchemaV3

water.bindings.pojos.ModelParametersSchemaV3

water.bindings.pojos.GLRMParametersV3

public class GLRMParametersV3
extends ModelParametersSchemaV3

Field Summary

Fields

Modifier and Type	Field and Description
boolean	expandUserY Expand categorical columns in user-specified initial Y
double	gammaX Regularization weight on X matrix
double	gammaY Regularization weight on Y matrix
boolean	imputeOriginal Reconstruct original training data by reversing transform
GenmodelalgosglrmGlrmInitialization	init Initialization mode
double	initStepSize Initial step size
int	k Rank of matrix approximation
java.lang.String	loadingName [Deprecated] Use representation_name instead.
GenmodelalgosglrmGlrmLoss	loss Numeric loss function
GenmodelalgosglrmGlrmLoss[]	lossByCol Loss function by column (override)
int[]	lossByColIdx Loss function by column index (override)
int	maxIterations Maximum number of iterations
int	maxUpdates Maximum number of updates, defaults to 2*max_iterations
double	minStepSize Minimum step size
GenmodelalgosglrmGlrmLoss	multiLoss Categorical loss function
int	period Length of period (only used with periodic loss function)
boolean	recoverSvd Recover singular values and eigenvectors of XY
GenmodelalgosglrmGlrmRegularizer	regularizationX Regularization function for X matrix
GenmodelalgosglrmGlrmRegularizer	regularizationY Regularization function for Y matrix
java.lang.String	representationName Frame key to save resulting X
long	seed RNG seed for initialization
SVDMethod	svdMethod Method for computing SVD during initialization (Caution: Randomized is currently experimental and unstable)
DataInfoTransformType	transform Transformation of training data
FrameKeyV3	userX User-specified initial X
FrameKeyV3	userY User-specified initial Y

Fields inherited from class water.bindings.pojos.ModelParametersSchemaV3

aucType, categoricalEncoding, checkpoint, customDistributionFunc, customMetricFunc, distribution, exportCheckpointsDir, foldAssignment, foldColumn, gainsliftBins, huberAlpha, ignoreConstCols, ignoredColumns, keepCrossValidationFoldAssignment, keepCrossValidationModels, keepCrossValidationPredictions, maxCategoricalLevels, maxRuntimeSecs, modelId, nfolds, offsetColumn, parallelizeCrossValidation, quantileAlpha, responseColumn, scoreEachIteration, stoppingMetric, stoppingRounds, stoppingTolerance, trainingFrame, tweediePower, validationFrame, weightsColumn

Constructor Summary

Constructors

Constructor and Description
GLRMParametersV3() Public constructor

Method Summary

Modifier and Type	Method and Description
java.lang.String	toString() Return the contents of this object as a JSON String.

Methods inherited from class java.lang.Object

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

Field Detail

transform

public DataInfoTransformType transform

Transformation of training data

k

public int k

Rank of matrix approximation

loss

public GenmodelalgosglrmGlrmLoss loss

Numeric loss function

multiLoss

@SerializedName(value="multi_loss")
public GenmodelalgosglrmGlrmLoss multiLoss

Categorical loss function

lossByCol

@SerializedName(value="loss_by_col")
public GenmodelalgosglrmGlrmLoss[] lossByCol

Loss function by column (override)

lossByColIdx

@SerializedName(value="loss_by_col_idx")
public int[] lossByColIdx

Loss function by column index (override)

period

public int period

Length of period (only used with periodic loss function)

regularizationX

@SerializedName(value="regularization_x")
public GenmodelalgosglrmGlrmRegularizer regularizationX

Regularization function for X matrix

regularizationY

@SerializedName(value="regularization_y")
public GenmodelalgosglrmGlrmRegularizer regularizationY

Regularization function for Y matrix

gammaX

@SerializedName(value="gamma_x")
public double gammaX

Regularization weight on X matrix

gammaY

@SerializedName(value="gamma_y")
public double gammaY

Regularization weight on Y matrix

maxIterations

@SerializedName(value="max_iterations")
public int maxIterations

Maximum number of iterations

maxUpdates

@SerializedName(value="max_updates")
public int maxUpdates

Maximum number of updates, defaults to 2*max_iterations

initStepSize

@SerializedName(value="init_step_size")
public double initStepSize

Initial step size

minStepSize

@SerializedName(value="min_step_size")
public double minStepSize

Minimum step size

seed

public long seed

RNG seed for initialization

init

public GenmodelalgosglrmGlrmInitialization init

Initialization mode

svdMethod

@SerializedName(value="svd_method")
public SVDMethod svdMethod

Method for computing SVD during initialization (Caution: Randomized is currently experimental and unstable)

userY

@SerializedName(value="user_y")
public FrameKeyV3 userY

User-specified initial Y

userX

@SerializedName(value="user_x")
public FrameKeyV3 userX

User-specified initial X

loadingName

@SerializedName(value="loading_name")
public java.lang.String loadingName

[Deprecated] Use representation_name instead. Frame key to save resulting X.

representationName

@SerializedName(value="representation_name")
public java.lang.String representationName

Frame key to save resulting X

expandUserY

@SerializedName(value="expand_user_y")
public boolean expandUserY

Expand categorical columns in user-specified initial Y

imputeOriginal

@SerializedName(value="impute_original")
public boolean imputeOriginal

Reconstruct original training data by reversing transform

recoverSvd

@SerializedName(value="recover_svd")
public boolean recoverSvd

Recover singular values and eigenvectors of XY

Constructor Detail

GLRMParametersV3

public GLRMParametersV3()

Public constructor

Method Detail

toString

public java.lang.String toString()

Return the contents of this object as a JSON String.

Overrides:

toString in class ModelParametersSchemaV3