Regression Loss Functions

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. How does mean squared error treat a residual?A. It averages the predicted values onlyB. It squares the residual before averagingC. It takes only the sign of the residualD. It discards all negative residuals

2. Which loss grows linearly with the magnitude of a large residual?A. Mean absolute errorB. Mean squared errorC. A squared residual without averagingD. The quadratic region of Huber loss

3. What does the threshold in Huber loss control?A. How many layers the network containsB. How often the test set is shuffledC. The number of output classesD. Where the loss changes from quadratic to linear growth

4. Two predictions have residuals of 2 and 4. What is their mean squared error?A. 6B. 20C. 10D. 3

5. A regression dataset contains occasional extreme measurement errors. Which change directly reduces their influence compared with MSE?A. Remove the validation setB. Try Huber loss and compare validation performanceC. Square the errors a second timeD. Select a model using training loss alone

Review this module from its first lesson →

Regression Loss Functions

Your result