Get the optinmal feature sets

get_bestFeatureSets(
  BMR_perf,
  selectedPATH,
  lrn = c("svm", "randomForest", "nnTrain"),
  window_size = 5,
  confidence_level = 0.05
)

Arguments

BMR_perf

a data.frame from getBMR_perf_tune()

selectedPATH

character, path to the selected features

lrn

character, mlr learner.id

window_size

numeric, width of the search window

confidence_level

numeric, confidence level

Value

a data.frame with the optimal model number of features and selected features per task.id and learner.id