Question 6
UnclassifiedRandom Undersampling (RUS) addresses class imbalance by:
Correct answer: B
Explanation
Random Undersampling (RUS) reduces class imbalance by discarding examples from the majority class until the class counts are closer. This matches the definition of undersampling: "removing majority-class samples to balance the classes."
Why each option is right or wrong
A. Generating new minority-class samples
B. Removing majority-class samples to balance the classes
Random undersampling works by decreasing the number of observations in the overrepresented class, not by generating new minority examples or changing class labels. In this question’s setup, the imbalance is resolved by dropping majority-class records until the class distribution is closer to parity, which is the defining mechanism of RUS.
C. Reweighting the loss
D. Using a different optimizer