ImpSampClassif function

WEighted Relevance-based Combination Strategy (WERCS) algorithm for imbalanced classification problems

This function handles imbalanced classification problems using the importance/relevance provided to re-sample the data set. The relevance is used to introduce replicas of the most important examples and to remove the least important examples. This function combines random over-sampling with random under-sampling which are applied in the problem classes according to the corresponding relevance.

WERCSClassif(form, dat, C.perc = "balance")

Arguments

• form: A formula describing the prediction problem
• dat: A data frame containing the original (unbalanced) data set
• C.perc: A list containing the percentage(s) of random under- or over-sampling to apply to each class. The over-sampling percentage is a number above 1 while the under-sampling percentage should be a number below 1. If the number 1 is provided for a given class then that class remains unchanged. Alternatively it may be "balance" (the default) or "extreme", cases where the sampling percentages are automatically estimated.

Returns

The function returns a data frame with the new data set resulting from the application of the importance sampling strategy.

Author(s)

Paula Branco paobranco@gmail.com , Rita Ribeiro rpribeiro@dcc.fc.up.pt and Luis Torgo ltorgo@dcc.fc.up.pt

See Also

RandUnderClassif, RandOverClassif

Examples

data(iris)
  # generating an artificially imbalanced data set
  ir <- iris[-c(51:70,111:150), ]
  IS.ext <-WERCSClassif(Species~., ir, C.perc = "extreme")
  IS.bal <-WERCSClassif(Species~., ir, C.perc = "balance")
  myIS <-WERCSClassif(Species~., ir, C.perc = list(setosa = 0.2,
                                                    versicolor = 2,
                                                    virginica = 6))
  # check the results
  table(ir$Species)
  table(IS.ext$Species)
  table(IS.bal$Species)
  table(myIS$Species)