A simple LearnerRegr which only analyses the response during train, ignoring all features. If hyperparameter robust is FALSE (default), constantly predicts mean(y) as response and sd(y) as standard error. If robust is TRUE, median() and madn() are used instead of mean() and sd(), respectively.

Format

R6::R6Class inheriting from LearnerRegr.

Construction

LearnerRegrFeatureless$new()
mlr_learners$get("regr.featureless")
lrn("regr.featureless")

See also

Dictionary of Learners: mlr_learners

as.data.table(mlr_learners) for a complete table of all (also dynamically created) Learner implementations.