01
Choose one
1 point
A housing dataset has one row for each home, with columns for size, neighborhood, and sale price. What is one sample in this dataset?
Free Practice Quiz Question List
2 Datasets and Features Online Quiz Questions
Use this free practice quiz with 20 questions to review 2 Datasets and Features, test your knowledge, and prepare for your next test or exam.
20 questions
02
True or false
1 point
A dataset can be accurate yet still be a poor choice if it does not reflect the real-world setting where the model will be used.
- A
True
- B
False
03
Fill in the blank
1 point
A model needs to use neighborhood names as input features. The categories may need to become usable by the model.
04
Written response
1 point
A housing dataset contains only homes from expensive neighborhoods, while a model will be used for homes across many neighborhoods. What data-quality property is most directly lacking?
05
Choose one
1 point
The exact outcome a model should predict is unavailable, so a team chooses a related measurable outcome as its label. Which conclusion about that label is most appropriate?
- A
Treat it as identical to the intended outcome because it is recorded in the dataset
- B
Recognize that it is an approximation that may not match the intended outcome
- C
Use it only as a feature and never as a prediction target
- D
Discard all samples that do not have the exact outcome
06
True or false
1 point
A team includes a feature that is recorded only after the event the model is supposed to predict. That feature is suitable for making the prediction because it is present in the dataset.
- A
True
- B
False
07
Choose all
1 point
A team is validating a housing dataset before modeling. Select all checks that belong in this inspection.
- A
Check whether rows and columns have the intended meanings
- B
Check types, units, plausible ranges, and missing values
- C
Choose the model architecture before inspecting the data
- D
Check for duplicate samples and inconsistent labels
08
Written response
1 point
After a team decides how to transform its model inputs, what data should receive those same transformations when the model is used?
09
Fill in the blank
1 point
When appropriate, a team may missing values using a suitable method.
10
Choose one
1 point
A team will evaluate a model using a test set and must scale a numerical feature. Which procedure avoids letting test data influence preprocessing?
- A
Use the entire dataset to learn scaling values before splitting it
- B
Learn scaling values from the test set so they match evaluation data
- C
Split into training and test sets, learn scaling values from training data, and apply them to test data
- D
Scale the training set but leave the test set in its original units
11
Choose all
1 point
A dataset has incorrect entries and missing values that cannot all be recovered. Select all appropriate ways to handle these problems during preparation.
- A
Correct erroneous values when possible
- B
Leave all errors unchanged so the model can learn from them
- C
If correction is not possible, consider removing affected samples
- D
If appropriate, fill missing values and record important preparation choices
12
Open ended
1 point
A housing dataset has neighborhood names as categories, missing size values, and inconsistent units. Describe how to prepare these data for prediction while preventing test-set leakage.
13
Choose one
1 point
During review, a team finds a home record whose sale price was copied from a different home. Which data-quality problem does this illustrate?
- A
A duplicate sample
- B
A missing feature value
- C
A category that needs encoding
- D
A mislabeled example
14
Choose one
1 point
In the housing example, which attribute is the label when the model is intended to predict a home's sale price?
- A
Size
- B
Sale price
- C
Neighborhood
- D
Number of bedrooms
15
Choose one
1 point
A team collects audio clips to build a model. Which statement best describes whether this collection can be a dataset?
- A
A dataset must have rows and columns, so an audio collection is not a dataset.
- B
Audio can be a dataset only if every recording has a numerical label.
- C
An audio collection can be a dataset even if it is not arranged as a table.
- D
Audio recordings are features, but cannot be samples.
16
Choose one
1 point
A team is validating a housing dataset before modeling. Which review is most appropriate?
- A
Check column meanings, types, units, plausible ranges, missing values, duplicates, and label consistency.
- B
Check only whether every column contains numbers.
- C
Remove all rows that differ from the most common row.
- D
Use the test set to decide which values and categories are acceptable.
17
Choose one
1 point
A dataset contains missing values and values known to be erroneous. Which preparation approach best follows sound data-handling practices?
- A
Replace all missing values with zero and leave erroneous values unchanged.
- B
Remove every sample containing a value outside the usual range without checking it.
- C
Ignore missing and erroneous values whenever a dataset is large.
- D
Correct erroneous values when possible; otherwise consider removing affected samples. Fill missing values with a suitable method when appropriate, and record important choices.
18
True or false
1 point
A model uses a property value that is recorded only after the sale, even though it is supposed to predict the sale price beforehand. This can make the model's results misleading.
- A
True
- B
False
19
Written response
1 point
A model cannot directly use neighborhood names in their current form, so the team converts the categories into numerical representations. What is this process called?
20
Written response
1 point
A team uses the test set to calculate the average that will fill missing values, then reports test performance. What evaluation problem can this cause?