Mainly for internal use. Trains a wrapped learner on a given training set. You have to implement this method if you want to add another learner to this package.

trainLearner(.learner, .task, .subset, .weights = NULL, ...)

Arguments

.learner

(RLearner)
Wrapped learner.

.task

(Task)
Task to train learner on.

.subset

(integer)
Subset of cases for training set, index the task with this. You probably want to use getTaskData for this purpose.

.weights

(numeric)
Weights for each observation.

...

(any)
Additional (hyper)parameters, which need to be passed to the underlying train function.

Value

(any). Model of the underlying learner.

Details

Your implementation must adhere to the following: The model must be fitted on the subset of .task given by .subset. All parameters in ... must be passed to the underlying training function.