Parameters of H2OGridSearch

Affected Class

  • ai.h2o.sparkling.ml.algos.H2OGridSearch

Parameters

  • Each parameter has also a corresponding getter and setter method. (E.g.: label -> getLabel() , setLabel(...) )

algo

Specifies the algorithm for grid search

Scala default value: null ; Python default value: None

hyperParameters

Hyper Parameters

Scala default value: Map() ; Python default value: {}

parallelism

Level of model-building parallelism, the possible values are:

  • 0 -> H2O selects parallelism level based on cluster configuration, such as number of cores

  • 1 -> Sequential model building, no parallelism

  • n>1 -> n models will be built in parallel if possible

Default value: 1

selectBestModelBy

Select best model by specific metric.If this value is not specified that the first model os taken.

Default value: "AUTO"

maxModels

Maximum number of models to build (optional).

Default value: 0

maxRuntimeSecs

Maximum time to spend building models (optional).

Default value: 0.0

seed

Seed for random number generator; set to a value other than -1 for reproducibility.

Scala default value: -1L ; Python default value: -1

stoppingMetric

Metric to use for early stopping (AUTO: logloss for classification, deviance for regression). Possible values are "AUTO", "deviance", "logloss", "MSE", "RMSE", "MAE", "RMSLE", "AUC", "AUCPR", "lift_top_group", "misclassification", "mean_per_class_error", "anomaly_score", "custom", "custom_increasing".

Default value: "AUTO"

stoppingRounds

Early stopping based on convergence of stopping_metric. Stop if simple moving average of length k of the stopping_metric does not improve for k:=stopping_rounds scoring events (0 to disable).

Default value: 0

stoppingTolerance

Relative tolerance for metric-based stopping criterion (stop if relative improvement is not at least this much).

Default value: 0.001

strategy

Hyperparameter space search strategy. Possible values are "Unknown", "Cartesian", "RandomDiscrete", "Sequential".

Default value: "Cartesian"