Oversample undersample

[Moelf]

It's missing from MLJ and using MLUtils.jl is sub optimal because they don't work with DataFrame which is very prevalent int MLJ.

[ablaom]

There is one package for DataFrames-only class imbalance, referenced here: https://julialang.org/jsoc/gsoc/MLJ/#correcting_for_class_imbalance_in_classification_problems . And there is a POC for integrating into MLJ here: #661 . And, I should add that MLUtils.jl has just recently added Tables compatability. So getobs should now work on DataFrames.jl ; see JuliaML/MLUtils.jl#124

[Moelf]

JuliaML/MLUtils.jl#129

[ablaom]

Closed by new packages Imbalance.jl and MLJBalancing.jl.