ML Foundations
Back to Evaluation, Leakage, and Unsupervised

Quiz: Spotting and Preventing Leakage

Practice quiz: Leakage is information from the future or from the target sneaking into training. The most common ML bug, the hardest to detect, the easiest to make.

6 minutesInteractive tutorial
🎯 Free Guest Mode: You are learning for free. Sign in to save your completion progress and quiz answers.
Question 1 of 8Answer saved0 answered

Your fraud model achieves AUC = 0.99 on the test set. What's your first hypothesis?