Model Fit and Data Scaling

Five questions. Choose one answer for each, then submit to see your score and explanations.

The practice target is 80%. You can retry as often as you like, and every lesson stays accessible.

You can practice as a guest. Sign in before starting an attempt to save your scores.

1. What commonly indicates underfitting?A. A validation set with no labelsB. A very low learning rate in every possible situationC. Poor performance on both training and validation dataD. Perfect training performance but poor validation performance only

2. Why fit a scaler on the training split only?A. To make every feature normally distributedB. To prevent information from held-out data influencing preprocessingC. To ensure all features become integersD. To force validation values inside the training range

3. What does z-score standardization subtract and divide by?A. The training mean and training standard deviationB. The training maximum and training minimumC. The test mean and test sample countD. The feature index and number of classes

4. A feature uses min–max scaling with minimum 10 and maximum 30. What does 15 become?A. 0.5B. 0.75C. 1.5D. 0.25

5. A task depends on very fine image edges. What is a risk of heavy smoothing?A. It guarantees better generalizationB. It removes the need to normalize inputsC. It can remove details needed for predictionD. It always increases spatial resolution

Review this module from its first lesson →

Model Fit and Data Scaling

Your result