A fraud dataset has only 1 percent positive examples. A data scientist wants the training procedure itself to place more emphasis on the minority class without discarding data. Which approach directly does that?
Answer & explanation
Correct answer: E
WHY E: Reweighting classes changes the training objective so minority-class mistakes matter more and directly mitigates imbalance. WHY NOT A: Feature dropping does not address class prevalence. WHY NOT B: Changing the test split does not fix training bias. WHY NOT D: Scaling may help optimization but not class imbalance. WHY NOT C: Row order alone does not rebalance the objective.