EasyDeepLearn

When should you use PR-AUC instead of ROC-AUC?

medium

Answer

  • Use PR-AUC when the positive class is rare (heavy imbalance).
  • ROC-AUC can look optimistic on imbalanced data because the true-negative rate dominates the false-positive rate.
  • PR-AUC only involves the positive class (precision and recall), so it reflects real performance on the minority class.
Check yourself — multiple choice
  • ROC-AUC is always the best metric
  • PR-AUC is preferred for highly imbalanced classes
  • Both are identical mathematically
  • PR-AUC ignores true positives

On rare-positive problems, ROC-AUC is misleading; PR-AUC focuses on the minority class.

#metrics#imbalance

Practise Supervised Learning

215 interview questions in this topic.

Related questions