Class Imbalance Handling Quiz

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

Quiz covering Data Integrity

What you'll be asked

  1. Which of the following is the primary risk of training a machine learning model on a highly imbalanced dataset without any adjustments?
  2. You are working on a fraud detection model where only 0.1% of transactions are fraudulent. Which metric is most appropriate for evaluating the model's performance?
  3. Which of the following techniques is considered a form of 'undersampling'?
  4. When using SMOTE (Synthetic Minority Over-sampling Technique) to address class imbalance, what is the core mechanism it uses to create new samples?
  5. You decide to use 'Class Weights' in your loss function instead of resampling. What is the main advantage of this approach?

Answers and explanations come after you submit.