ModelSelectionParametersV3 (h2o-bindings 3.44.0 API)

java.lang.Object
- water.bindings.pojos.SchemaV3
- - water.bindings.pojos.ModelParametersSchemaV3
  - - water.bindings.pojos.ModelSelectionParametersV3

public class ModelSelectionParametersV3
extends ModelParametersSchemaV3

Field Summary

Fields
Modifier and Type	Field and Description
`double[]`	`alpha` Distribution of regularization between the L1 (Lasso) and L2 (Ridge) penalties.
`boolean`	`balanceClasses` Balance training data class counts via over/under-sampling (for imbalanced data).
`FrameKeyV3`	`betaConstraints` Beta constraints
`double`	`betaEpsilon` Converge if beta changes less (using L-infinity norm) than beta esilon, ONLY applies to IRLSM solver
`boolean`	`buildGlmModel` For maxrsweep mode only.
`boolean`	`calcLike` if true, will return likelihood function value for HGLM.
`float[]`	`classSamplingFactors` Desired over/under-sampling ratios per class (in lexicographic order).
`boolean`	`coldStart` Only applicable to multiple alpha/lambda values.
`boolean`	`computePValues` Request p-values computation, p-values work only with IRLSM solver and no regularization
`boolean`	`earlyStopping` Stop early when there is no more relative improvement on train or validation (if provided)
`GLMFamily`	`family` Family.
`double`	`gradientEpsilon` Converge if objective changes less (using L-infinity norm) than this, ONLY applies to L-BFGS solver.
`GLMInfluence`	`influence` If set to dfbetas will calculate the difference in beta when a datarow is included and excluded in the dataset.
`boolean`	`intercept` Include constant term in the model
`double[]`	`lambda` Regularization strength
`double`	`lambdaMinRatio` Minimum lambda used in lambda search, specified as a ratio of lambda_max (the smallest lambda that drives all coefficients to zero).
`boolean`	`lambdaSearch` Use lambda search starting at lambda max, given lambda is then interpreted as lambda min
`GLMLink`	`link` Link function.
`int`	`maxActivePredictors` Maximum number of active predictors during computation.
`float`	`maxAfterBalanceSize` Maximum relative size of the training data after balancing class counts (can be less than 1.0).
`int`	`maxConfusionMatrixSize` [Deprecated] Maximum size (# classes) for confusion matrices to be printed in the Logs
`int`	`maxIterations` Maximum number of iterations
`int`	`maxPredictorNumber` Maximum number of predictors to be considered when building GLM models.
`int`	`minPredictorNumber` For mode = 'backward' only.
`GLMMissingValuesHandling`	`missingValuesHandling` Handling of missing values.
`ModelSelectionMode`	`mode` Mode: Used to choose model selection algorithms to use.
`boolean`	`multinodeMode` For maxrsweep only.
`int`	`nlambdas` Number of lambdas to be used in a search.
`boolean`	`nonNegative` Restrict coefficients (not intercept) to be non-negative
`int`	`nparallelism` number of models to build in parallel.
`double`	`objectiveEpsilon` Converge if objective value changes less than this.
`double`	`objReg` Likelihood divider in objective value computation, default (of -1.0) will set it to 1/nobs
`FrameKeyV3`	`plugValues` Plug Values (a single row frame containing values that will be used to impute missing values of the training/validation frame, use with conjunction missing_values_handling = PlugValues)
`double`	`prior` Prior probability for y==1.
`double`	`pValuesThreshold` For mode='backward' only.
`boolean`	`removeCollinearColumns` In case of linearly dependent columns, remove some of the dependent columns
`int`	`scoreIterationInterval` Perform scoring for every score_iteration_interval iterations
`long`	`seed` Seed for pseudo random number generator (if applicable)
`GLMSolver`	`solver` AUTO will set the solver based on given data and the other parameters.
`boolean`	`standardize` Standardize numeric columns to have zero mean and unit variance
`double[]`	`startval` double array to initialize fixed and random coefficients for HGLM, coefficients for GLM.
`double`	`theta` Theta
`double`	`tweedieLinkPower` Tweedie link power
`double`	`tweedieVariancePower` Tweedie variance power

Constructor Summary

Constructors
Constructor and Description

ModelSelectionParametersV3()
Public constructor

Constructors
Constructor and Description
`ModelSelectionParametersV3()` Public constructor

Method Summary

All Methods Instance Methods Concrete Methods
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

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

- Field Detail
  - seed
```
public long seed
```
    Seed for pseudo random number generator (if applicable)
  - family
```
public GLMFamily family
```
    Family. For maxr/maxrsweep, only gaussian. For backward, ordinal and multinomial families are not supported
  - tweedieVariancePower
```
@SerializedName(value="tweedie_variance_power")
public double tweedieVariancePower
```
    Tweedie variance power
  - tweedieLinkPower
```
@SerializedName(value="tweedie_link_power")
public double tweedieLinkPower
```
    Tweedie link power
  - theta
```
public double theta
```
    Theta
  - solver
```
public GLMSolver solver
```
    AUTO will set the solver based on given data and the other parameters. IRLSM is fast on on problems with small number of predictors and for lambda-search with L1 penalty, L_BFGS scales better for datasets with many columns.
  - alpha
```
public double[] alpha
```
    Distribution of regularization between the L1 (Lasso) and L2 (Ridge) penalties. A value of 1 for alpha represents Lasso regression, a value of 0 produces Ridge regression, and anything in between specifies the amount of mixing between the two. Default value of alpha is 0 when SOLVER = 'L-BFGS'; 0.5 otherwise.
  - lambda
```
public double[] lambda
```
    Regularization strength
  - lambdaSearch
```
@SerializedName(value="lambda_search")
public boolean lambdaSearch
```
    Use lambda search starting at lambda max, given lambda is then interpreted as lambda min
  - multinodeMode
```
@SerializedName(value="multinode_mode")
public boolean multinodeMode
```
    For maxrsweep only. If enabled, will attempt to perform sweeping action using multiple nodes in the cluster. Defaults to false.
  - buildGlmModel
```
@SerializedName(value="build_glm_model")
public boolean buildGlmModel
```
    For maxrsweep mode only. If true, will return full blown GLM models with the desired predictorsubsets. If false, only the predictor subsets, predictor coefficients are returned. This is forspeeding up the model selection process. The users can choose to build the GLM models themselvesby using the predictor subsets themselves. Defaults to false.
  - earlyStopping
```
@SerializedName(value="early_stopping")
public boolean earlyStopping
```
    Stop early when there is no more relative improvement on train or validation (if provided)
  - nlambdas
```
public int nlambdas
```
    Number of lambdas to be used in a search. Default indicates: If alpha is zero, with lambda search set to True, the value of nlamdas is set to 30 (fewer lambdas are needed for ridge regression) otherwise it is set to 100.
  - scoreIterationInterval
```
@SerializedName(value="score_iteration_interval")
public int scoreIterationInterval
```
    Perform scoring for every score_iteration_interval iterations
  - standardize
```
public boolean standardize
```
    Standardize numeric columns to have zero mean and unit variance
  - coldStart
```
@SerializedName(value="cold_start")
public boolean coldStart
```
    Only applicable to multiple alpha/lambda values. If false, build the next model for next set of alpha/lambda values starting from the values provided by current model. If true will start GLM model from scratch.
  - missingValuesHandling
```
@SerializedName(value="missing_values_handling")
public GLMMissingValuesHandling missingValuesHandling
```
    Handling of missing values. Either MeanImputation, Skip or PlugValues.
  - plugValues
```
@SerializedName(value="plug_values")
public FrameKeyV3 plugValues
```
    Plug Values (a single row frame containing values that will be used to impute missing values of the training/validation frame, use with conjunction missing_values_handling = PlugValues)
  - nonNegative
```
@SerializedName(value="non_negative")
public boolean nonNegative
```
    Restrict coefficients (not intercept) to be non-negative
  - maxIterations
```
@SerializedName(value="max_iterations")
public int maxIterations
```
    Maximum number of iterations
  - betaEpsilon
```
@SerializedName(value="beta_epsilon")
public double betaEpsilon
```
    Converge if beta changes less (using L-infinity norm) than beta esilon, ONLY applies to IRLSM solver
  - objectiveEpsilon
```
@SerializedName(value="objective_epsilon")
public double objectiveEpsilon
```
    Converge if objective value changes less than this. Default (of -1.0) indicates: If lambda_search is set to True the value of objective_epsilon is set to .0001. If the lambda_search is set to False and lambda is equal to zero, the value of objective_epsilon is set to .000001, for any other value of lambda the default value of objective_epsilon is set to .0001.
  - gradientEpsilon
```
@SerializedName(value="gradient_epsilon")
public double gradientEpsilon
```
    Converge if objective changes less (using L-infinity norm) than this, ONLY applies to L-BFGS solver. Default (of -1.0) indicates: If lambda_search is set to False and lambda is equal to zero, the default value of gradient_epsilon is equal to .000001, otherwise the default value is .0001. If lambda_search is set to True, the conditional values above are 1E-8 and 1E-6 respectively.
  - objReg
```
@SerializedName(value="obj_reg")
public double objReg
```
    Likelihood divider in objective value computation, default (of -1.0) will set it to 1/nobs
  - link
```
public GLMLink link
```
    Link function.
  - startval
```
public double[] startval
```
    double array to initialize fixed and random coefficients for HGLM, coefficients for GLM.
  - calcLike
```
@SerializedName(value="calc_like")
public boolean calcLike
```
    if true, will return likelihood function value for HGLM.
  - mode
```
public ModelSelectionMode mode
```
    Mode: Used to choose model selection algorithms to use. Options include 'allsubsets' for all subsets, 'maxr' that uses sequential replacement and GLM to build all models, slow but works with cross-validation, validation frames for more robust results, 'maxrsweep' that uses sequential replacement and sweeping action, much faster than 'maxr', 'backward' for backward selection.
  - intercept
```
public boolean intercept
```
    Include constant term in the model
  - prior
```
public double prior
```
    Prior probability for y==1. To be used only for logistic regression iff the data has been sampled and the mean of response does not reflect reality.
  - lambdaMinRatio
```
@SerializedName(value="lambda_min_ratio")
public double lambdaMinRatio
```
    Minimum lambda used in lambda search, specified as a ratio of lambda_max (the smallest lambda that drives all coefficients to zero). Default indicates: if the number of observations is greater than the number of variables, then lambda_min_ratio is set to 0.0001; if the number of observations is less than the number of variables, then lambda_min_ratio is set to 0.01.
  - betaConstraints
```
@SerializedName(value="beta_constraints")
public FrameKeyV3 betaConstraints
```
    Beta constraints
  - maxActivePredictors
```
@SerializedName(value="max_active_predictors")
public int maxActivePredictors
```
    Maximum number of active predictors during computation. Use as a stopping criterion to prevent expensive model building with many predictors. Default indicates: If the IRLSM solver is used, the value of max_active_predictors is set to 5000 otherwise it is set to 100000000.
  - balanceClasses
```
@SerializedName(value="balance_classes")
public boolean balanceClasses
```
    Balance training data class counts via over/under-sampling (for imbalanced data).
  - classSamplingFactors
```
@SerializedName(value="class_sampling_factors")
public float[] classSamplingFactors
```
    Desired over/under-sampling ratios per class (in lexicographic order). If not specified, sampling factors will be automatically computed to obtain class balance during training. Requires balance_classes.
  - maxAfterBalanceSize
```
@SerializedName(value="max_after_balance_size")
public float maxAfterBalanceSize
```
    Maximum relative size of the training data after balancing class counts (can be less than 1.0). Requires balance_classes.
  - maxConfusionMatrixSize
```
@SerializedName(value="max_confusion_matrix_size")
public int maxConfusionMatrixSize
```
    [Deprecated] Maximum size (# classes) for confusion matrices to be printed in the Logs
  - computePValues
```
@SerializedName(value="compute_p_values")
public boolean computePValues
```
    Request p-values computation, p-values work only with IRLSM solver and no regularization
  - removeCollinearColumns
```
@SerializedName(value="remove_collinear_columns")
public boolean removeCollinearColumns
```
    In case of linearly dependent columns, remove some of the dependent columns
  - maxPredictorNumber
```
@SerializedName(value="max_predictor_number")
public int maxPredictorNumber
```
    Maximum number of predictors to be considered when building GLM models. Defaults to 1.
  - minPredictorNumber
```
@SerializedName(value="min_predictor_number")
public int minPredictorNumber
```
    For mode = 'backward' only. Minimum number of predictors to be considered when building GLM models starting with all predictors to be included. Defaults to 1.
  - nparallelism
```
public int nparallelism
```
    number of models to build in parallel. Defaults to 0.0 which is adaptive to the system capability
  - pValuesThreshold
```
@SerializedName(value="p_values_threshold")
public double pValuesThreshold
```
    For mode='backward' only. If specified, will stop the model building process when all coefficientsp-values drop below this threshold
  - influence
```
public GLMInfluence influence
```
    If set to dfbetas will calculate the difference in beta when a datarow is included and excluded in the dataset.
- Constructor Detail
  - ModelSelectionParametersV3
```
public ModelSelectionParametersV3()
```
    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

Class ModelSelectionParametersV3

Field Summary

Fields inherited from class water.bindings.pojos.ModelParametersSchemaV3

Constructor Summary

Method Summary

Methods inherited from class java.lang.Object

Field Detail

seed

family

tweedieVariancePower

tweedieLinkPower

theta

solver

alpha

lambda

lambdaSearch

multinodeMode

buildGlmModel

earlyStopping

nlambdas

scoreIterationInterval

standardize

coldStart

missingValuesHandling

plugValues

nonNegative

maxIterations

betaEpsilon

objectiveEpsilon

gradientEpsilon

objReg

link

startval

calcLike

mode

intercept

prior

lambdaMinRatio

betaConstraints

maxActivePredictors

balanceClasses

classSamplingFactors

maxAfterBalanceSize

maxConfusionMatrixSize

computePValues

removeCollinearColumns

maxPredictorNumber

minPredictorNumber

nparallelism

pValuesThreshold

influence

Constructor Detail

ModelSelectionParametersV3

Method Detail

toString