01
Choose one
1 point
A supervised model uses a home's size and location to predict its sale price. Which item is the label?
Free Practice Quiz Question List
2 Features, Labels, and Data Preparation Online Quiz Questions
Use this free practice quiz with 20 questions to review 2 Features, Labels, and Data Preparation, test your knowledge, and prepare for your next test or exam.
20 questions
02
Fill in the blank
1 point
Use the to fit the model; use the to compare models during development.
03
Written response
1 point
A model needs numeric input, so a category such as "blue" is converted into columns that indicate category membership. What are these columns called?
04
Choose one
1 point
A team is comparing model settings before choosing a final model. Which dataset should it use to guide that choice?
- A
Use it to make the final evaluation after all development choices are complete.
- B
Use its results during development to compare models and select settings.
- C
Use it only to fit the model's parameters.
05
Choose all
1 point
Before training a supervised model, which data-quality checks are appropriate? Choose all that apply.
- A
Check that each example is paired with its correct label.
- B
Remove every row with any missing value without considering the situation.
- C
Look for missing, invalid, or inconsistent values.
- D
Assume duplicate examples are harmless and do not check for them.
- E
Look for duplicate examples.
06
Fill in the blank
1 point
A can help keep preprocessing consistent and prevent .
07
Choose one
1 point
A team is building a model to predict appointment cancellations before appointments occur. Which feature choice best fits the requirements for a useful prediction feature?
- A
Use the final cancellation status as an input, even though it is recorded after the appointment.
- B
Use a post-appointment survey response to predict whether the appointment will be missed.
- C
Use information available before the appointment that is useful for predicting cancellation.
08
True or false
1 point
True or false: For time-dependent data, splitting chronologically so training examples precede validation and test examples can better reflect how a model will encounter future data.
- A
True
- B
False
09
Written response
1 point
A dataset is split by reserving 20% for testing first. A second split must take 15% of the original dataset from the remaining data for validation. What test_size value should the second split use?
10
Choose one
1 point
A model will be used to forecast future demand from historical observations. Which split design best reflects that use?
- A
Randomly distribute all dates across all subsets.
- B
Put the latest examples in training and the earliest examples in testing.
- C
Place earlier examples in training, followed by validation and test examples in time order.
11
Choose all
1 point
A dataset has enough examples in each class, includes multiple rows per household, and represents time-dependent observations. Which splitting practices match these conditions? Choose all that apply.
- A
Use stratification when preserving class proportions matters and each class has enough examples.
- B
Put rows from the same household into different subsets to increase the number of examples in each one.
- C
Keep all rows belonging to one household in a single subset.
- D
For time-dependent data, split in chronological order.
- E
Use stratification to ensure that training examples come earlier in time than test examples.
12
True or false
1 point
If a team repeatedly consults test-set results while tuning a model, those results may influence development choices and weaken the test set's role as an independent final check.
- A
True
- B
False
13
Open ended
1 point
A supervised dataset contains missing values in a numeric feature, and the team plans to scale that feature. Describe a safe workflow from splitting the data through final evaluation. Include where imputation and scaling are fitted, how the fitted transformations are used on other data, and how validation and test results should be used.
14
Choose one
1 point
A supervised-learning model uses a home’s size and location to predict its sale price. Which value is the label?
- A
The home’s size
- B
The home’s sale price
- C
The home’s location
- D
The model’s prediction error
15
Choose one
1 point
A team is comparing model settings before choosing a final model. Which dataset subset should it use to guide that development decision?
- A
To fit the final model and learn its patterns
- B
To provide the final evaluation after all development choices
- C
To compare models and guide development choices
- D
To replace the need for a test set
16
True or false
1 point
True or false: Repeatedly consulting test-set results while tuning a model can weaken the test set’s role as an independent final evaluation.
- A
True
- B
False
17
Choose one
1 point
During a data check, a team finds that some rows may be paired with the wrong outcomes. What should it verify before training?
- A
Verify that each example is paired with its correct label
- B
Remove every example whose features are numeric
- C
Use the test set to correct labels during model tuning
- D
Keep duplicate rows so that each example appears more often
18
Choose one
1 point
A model requires numeric inputs, but one feature contains categories such as “blue.” Which preparation step is consistent with the material?
- A
Treat “blue” as the label in every prediction task
- B
Represent categories such as “blue” with indicator columns
- C
Remove all categorical features before examining the data
- D
Use the category to determine the test-set size
19
Written response
1 point
A dataset is split by first reserving 20% for testing, then reserving 0.1875 of the remaining data for validation. What percentage of the original dataset goes to validation? Enter a numeric percentage without the percent sign.
20
Written response
1 point
A dataset includes calendar values as inputs, but the model needs numeric values. Which feature data type might be converted into useful components during preparation?