Undersampling is a data cleaning technique for dealing with imbalanced data in machine learning. Imbalanced data means that some classes or labels are overrepresented or underrepresented in the ...