SL.xgboost.bounded.Rd
A SuperLearner wrapper for SL.xgboost that ensures positive predictions. Predictions below 1e-4 are clipped to a minimum value of 1e-4.
SL.xgboost.bounded(..., ntrees = 100, lower = 1e-04)
SL.xgboost arguments.
Number of trees, default is 100.
Lower bound to clip predictions to, default is 1e-4.
A list with components:
pred: A numeric vector of predictions.
pred
fit: A list containing the fitted model object.
fit