makeWrappedLearner: Build a wrapped learner using a specified learning strategy.

In Prometheus77/ezmlr: Easy Interface to mlr

Description

Follows a specified learning strategy (Strategy) to create a wrapped learner customized to a specific task. May include things such as:

• Preprocessing

• Imputation of missing values

• Feature selection

• Feature filtering

• Upsampling/downsampling

• Hyperparameter tuning

Usage

makeWrappedLearner(strategy, id = strategy, task, strategy.pars = list(),
  learner.pars = list(), ..., predict.type = NULL,
  predict.threshold = NULL, fix.factors.prediction = TRUE, config = NULL)

Arguments

strategy	(character(1)) Name (short name) of the strategy to use for training.
id	(character(1)) Learner ID. Default is strategy.
task	(Task) Task to which the strategy should be applied.
strategy.pars	(list) Optional parameters to pass to the strategy. These will vary strategy by strategy.
learner.pars	(list) Optional parameters to pass to the wrapped learner. These are the standard parameters used by each learner in mlr.

Prometheus77/ezmlr documentation built on May 17, 2019, 6:34 p.m.