← All types
AI & MLOb Object

MLHyperparameters

XGBoost hyperparameters for model training

Fields

maxDepth

IntNullable

Maximum tree depth (default 6)

eta

FloatNullable

Learning rate (default 0.3)

rounds

IntNullable

Number of boosting rounds (default 100)

Row subsampling ratio (default 1.0)

Column subsampling ratio (default 1.0)

Minimum sum of instance weight in a child

gamma

FloatNullable

Minimum loss reduction for split

Evaluation metric override

Referenced by (2)
Exact type SDL
"""XGBoost hyperparameters for model training"""
type MLHyperparameters {
  """Maximum tree depth (default 6)"""
  maxDepth: Int
  """Learning rate (default 0.3)"""
  eta: Float
  """Number of boosting rounds (default 100)"""
  rounds: Int
  """Row subsampling ratio (default 1.0)"""
  subsample: Float
  """Column subsampling ratio (default 1.0)"""
  colsampleBytree: Float
  """Minimum sum of instance weight in a child"""
  minChildWeight: Float
  """Minimum loss reduction for split"""
  gamma: Float
  """Evaluation metric override"""
  evalMetric: String
}