oversample or undersample before every model fit. Note that observation weights do not influence the sampling and are simply passed
down to the next learner.makeUndersampleWrapper(learner, usw.rate = 1, usw.cl = NULL)makeOversampleWrapper(learner, osw.rate = 1, osw.cl = NULL)
Learner | character(1)]
The learner.
If you pass a string the learner will be created via makeLearner.numeric(1)]
Factor to downsample a class. Must be between 0 and 1,
where 1 means no downsampling, 0.5 implies reduction to 50 percent
and 0 would imply reduction to 0 observations.
Default is 1.character(1)]
Class that should be undersampled.
Default is NULL, which means the larger one.numeric(1)]
Factor to oversample a class. Must be between 1 and Inf,
where 1 means no oversampling and 2 would mean doubling the class size.
Default is 1.character(1)]
Class that should be oversampled.
Default is NULL, which means the smaller one.Learner].makeOverBaggingWrapper,
oversample, smote Other wrapper: makeBaggingWrapper,
makeConstantClassWrapper,
makeCostSensClassifWrapper,
makeCostSensRegrWrapper,
makeDownsampleWrapper,
makeFeatSelWrapper,
makeFilterWrapper,
makeImputeWrapper,
makeMulticlassWrapper,
makeMultilabelBinaryRelevanceWrapper,
makeMultilabelClassifierChainsWrapper,
makeMultilabelDBRWrapper,
makeMultilabelNestedStackingWrapper,
makeMultilabelStackingWrapper,
makeOverBaggingWrapper,
makePreprocWrapperCaret,
makePreprocWrapper,
makeRemoveConstantFeaturesWrapper,
makeSMOTEWrapper,
makeTuneWrapper,
makeWeightedClassesWrapper