Timedora
Free Practice Quiz Question List

4 Regression Online Quiz Questions

Use this free practice quiz with 20 questions to review 4 Regression, test your knowledge, and prepare for your next test or exam.

20 questions
01
Choose one
1 point

Which task is a regression problem?

  1. A

    Predicting a home’s sale price from its features

  2. B

    Assigning an email to a spam or not-spam category

  3. C

    Grouping customers without a known target

02
Fill in the blank
1 point

Lasso regularization can shrink some coefficients to .

03
True or false
1 point

True or false: If a linear regression coefficient says predicted rent rises with floor area, that alone proves that increasing floor area causes rent to rise.

  1. A

    True

  2. B

    False

04
Choose one
1 point

How does a mean baseline make predictions in a regression task?

  1. A

    It predicts a separate average for every feature combination.

  2. B

    It predicts the average target observed in training for every example.

  3. C

    It divides examples into groups and predicts a value for each group.

05
Written response
1 point

For the model rent^=700+18(area in m2)\widehat{\text{rent}} = 700 + 18(\text{area in m}^2), what is the predicted change in rent for each additional square meter? Enter the numerical coefficient only.

06
Fill in the blank
1 point

A regression model predicts 4747 when the actual target is 5252. Using residual =y−y^= y - \hat{y}, the residual is .

07
Choose one
1 point

What objective does ordinary least squares use to choose coefficients in linear regression?

  1. A

    The largest absolute prediction error

  2. B

    The number of input features

  3. C

    The sum of squared differences between observed and predicted targets

08
Choose all
1 point

Select all statements that correctly describe regularized linear regression.

  1. A

    Ridge and Lasso add penalties on coefficient size.

  2. B

    Regularization can help when features overlap in the information they provide.

  3. C

    Lasso can shrink some coefficients to zero.

  4. D

    Regularization guarantees better performance on every unseen example.

09
Written response
1 point

What is the term for the difference y−y^y - \hat{y} between an actual target and a model’s prediction?

10
Choose all
1 point

Select all statements that accurately describe ensemble regression methods.

  1. A

    A random forest averages predictions from many trees.

  2. B

    Gradient-boosted trees build trees in sequence to improve predictions.

  3. C

    Ensemble regression can capture complex relationships.

  4. D

    Both methods are always easier to interpret than a small linear model.

11
True or false
1 point

True or false: A linear model’s prediction far outside the range of its training data is necessarily realistic because the model has a linear equation.

  1. A

    True

  2. B

    False

12
Choose one
1 point

Which approach provides evidence about how well a regression model generalizes to new examples?

  1. A

    Evaluate the model only on the training examples and treat that score as proof of generalization.

  2. B

    Compare its predictions with the training targets after fitting.

  3. C

    Assess its performance on held-out data not used to train it.

13
Open ended
1 point

A model was trained on homes with areas from 40 m240\,\mathrm{m}^2 to 160 m2160\,\mathrm{m}^2, but it predicts rent for a 500 m2500\,\mathrm{m}^2 home. Explain why this prediction deserves caution and identify at least one check that could help assess its reliability.

14
Choose one
1 point

A team wants to predict the number of minutes each delivery will take. Which statement best explains why this is a regression task?

  1. A

    Predicting whether a delivery will be early, on time, or late

  2. B

    Predicting how many minutes a delivery will take

  3. C

    Assigning each delivery to a customer segment

  4. D

    Identifying which delivery method was used

15
Choose one
1 point

A team uses a mean baseline for a regression problem. What does this model predict for each example?

  1. A

    The most frequent target from the training data

  2. B

    The target value of the closest training example

  3. C

    The average target in the training data

  4. D

    A different randomly chosen target for each example

16
Choose one
1 point

Which statement correctly distinguishes random forests from gradient-boosted trees in regression?

  1. A

    Both methods use only one tree, but with different feature sets.

  2. B

    Random forests average predictions from many trees; gradient boosting builds trees in sequence.

  3. C

    Random forests build trees in sequence; gradient boosting averages independent tree predictions.

  4. D

    Both methods predict the average target from the training data.

17
Choose one
1 point

A regression model has many overlapping features. Which statement about Ridge and Lasso is supported by the material?

  1. A

    They ensure every feature has a nonzero coefficient.

  2. B

    They make a model predict the same value for every example.

  3. C

    They penalize coefficient size, and Lasso can shrink some coefficients to zero.

  4. D

    They remove the need to assess a model on unseen data.

18
True or false
1 point

True or false: A regression model's prediction is guaranteed to be reliable even when important input features are missing.

  1. A

    True

  2. B

    False

19
Written response
1 point

A simplified rent model is rent^=700+18(area in m2)\widehat{\text{rent}} = 700 + 18(\text{area in }\mathrm{m}^2). What rent does it predict for an area of 10 m210\,\mathrm{m}^2?

20
Written response
1 point

What is the regression term for a prediction made far outside the range of the training data?