Type Members

1.  class MultiTaskTrainingNode extends AnyRef

   Node in a multi-task training tree, which can produce nodes for its model trees.

   Node in a multi-task training tree, which can produce nodes for its model trees. Splits are chosen using a MultiTaskSplitter, which considers the sum impurity decrease across all tasks.

2.  case class MultiTaskTreeLearner(numFeatures: Int = -1, maxDepth: Int = 30, minLeafInstances: Int = 1, randomizePivotLocation: Boolean = false, rng: Random = Random) extends MultiTaskLearner with Product with Serializable

   A tree learner that operates on multiple labels.

   A tree learner that operates on multiple labels.

   numFeatures

   to random select from at each split (numbers less than 0 indicate that all features are used)

   maxDepth

   to grow the tree to

   minLeafInstances

   minimum number of training instances per leaf

   randomizePivotLocation

   whether to generate splits randomly between the data points

   rng

   random number generator, for reproducibility

3.  class MultiTaskTreeTrainingResult extends MultiTaskTrainingResult