dynamic_resamplebm: Dynamic resampling using learners from benchmarking results

In messamat/globalIRmap: What the Package Does (One Line, Title Case)

Description

Run a train-predict-test routine using a learner from a BenchmarkResult: Repeatedly apply learner on a training set of task to train a model, then use the trained model to predict observations of a test set. Training and test sets are defined by the resampling.

Usage

dynamic_resamplebm(
  in_task,
  in_bm,
  in_lrnid,
  in_resampling,
  type,
  inp_resdir = NULL,
  store_models = FALSE
)

Arguments

in_task	Task to resample
in_bm	BenchmarkResult from which to extract learner or name of serialized file (.qs) on disk (accessed through qread)
in_resampling	Resampling cross-validation strategy
type	(character) Type of learner 'classif' or 'regr'. Other values are not accepted.
inp_resdir	(character) path to where qs file is located (excluding the name of the qs file)
store_models	whether to keep the fitted model after the test set has been predicted. Set to TRUE if you want to further analyse the models or want to extract information like variable importance.
in__lrnid	id of learner to extract from BenchmarkResult (e.g., "oversample.classif.ranger")

Details

The dynamic aspect of this model is that it runs 'reset_tuning' on the fly to make sure that the hyperparameter search space matches the task (e.g., if the number of candidate predictor variables has been reduced, it adjusts mtry)

Value

ResampleResult

messamat/globalIRmap documentation built on July 4, 2021, 10:48 a.m.