Confusion Matrix and ROC Curves Quiz

5 questions Pass: 70% +25 pts (50 with Pro)

Quiz covering Model Evaluation and Selection

What you'll be asked

  1. In a binary classification confusion matrix, what does the 'False Negative' (FN) value represent?
  2. You are evaluating a credit card fraud detection model on Azure. Since fraud is rare, the dataset is highly imbalanced. Which metric should you prioritize to assess performance?
  3. What does the Area Under the ROC Curve (AUC-ROC) score of 0.5 indicate?
  4. If you increase the probability threshold for a binary classifier, how does it typically affect the Precision and Recall?
  5. When comparing two models using ROC curves, Model A has a higher AUC than Model B, but Model B has a higher True Positive Rate at a very low False Positive Rate. Which model should you choose if your business requirement is to ensure almost zero false alarms, even if it means missing some positive cases?

Answers and explanations come after you submit.