Practice Library
All MCQs
Browse exam-wise, subject-wise, and country-wise MCQs with explanations.
Choose an option to check your answer.
A.
Distance calculations are otherwise dominated by large-scale features
B.
KNN requires binary attributes only
C.
Standardization creates labels
D.
KNN cannot handle positive values
Show Answer
Correct Answer: A. Distance calculations are otherwise dominated by large-scale features
Explanation:
KNN bases predictions on proximity in feature space.
Choose an option to check your answer.
A.
Values between zero and one in every sample
B.
Only positive integer values
C.
Equal category frequencies
D.
Values centered at zero with unit standard deviation
Show Answer
Correct Answer: D. Values centered at zero with unit standard deviation
Explanation:
Each value is transformed by subtracting the mean and dividing by the standard deviation.
Choose an option to check your answer.
A.
Centers values at the median only
B.
Makes the variance exactly one without shifting
C.
Maps values into a specified interval such as 0 to 1
D.
Converts numeric values to ranks
Show Answer
Correct Answer: C. Maps values into a specified interval such as 0 to 1
Explanation:
Min-max scaling uses the observed minimum and maximum.
Choose an option to check your answer.
A.
Converting all variables to categories
B.
Rescaling numeric attributes to a common range or scale
C.
Removing every duplicate row
D.
Replacing missing values with labels
Show Answer
Correct Answer: B. Rescaling numeric attributes to a common range or scale
Explanation:
Normalization prevents large-scale variables from dominating scale-sensitive algorithms.
Choose an option to check your answer.
A.
Obtaining a smaller representation that preserves important information
B.
Deleting all rare classes
C.
Converting every value to zero
D.
Increasing the number of dimensions
Show Answer
Correct Answer: A. Obtaining a smaller representation that preserves important information
Explanation:
Reduction can lower storage, computation, and noise.
Choose an option to check your answer.
A.
A shortage of observations
B.
A model with low variance
C.
A variable with a true zero
D.
Unnecessary duplication of information across attributes or records
Show Answer
Correct Answer: D. Unnecessary duplication of information across attributes or records
Explanation:
Redundancy can waste storage and overweight repeated information.
Choose an option to check your answer.
A.
Estimating a regression coefficient
B.
Assigning cluster centroids
C.
Determining which records refer to the same real-world entity
D.
Scaling variables to unit variance
Show Answer
Correct Answer: C. Determining which records refer to the same real-world entity
Explanation:
Names, addresses, and identifiers may vary across sources.
Choose an option to check your answer.
A.
Removing all duplicate features
B.
Combining data from multiple sources into a consistent dataset
C.
Sorting one table by a key
D.
Training several classifiers
Show Answer
Correct Answer: B. Combining data from multiple sources into a consistent dataset
Explanation:
Integration reconciles schemas, identifiers, formats, and overlapping records.
Choose an option to check your answer.
A.
Use the most frequent category or a separate 'Unknown' category
B.
Use the arithmetic mean
C.
Use the standard deviation
D.
Apply logarithmic transformation
Show Answer
Correct Answer: A. Use the most frequent category or a separate 'Unknown' category
Explanation:
Categorical values require a valid category rather than a numeric summary.
Choose an option to check your answer.
A.
The median always preserves correlations
B.
The median creates new observations
C.
The median guarantees normality
D.
The median is less affected by extreme values
Show Answer
Correct Answer: D. The median is less affected by extreme values
Explanation:
For skewed numeric variables, the median often better represents the center.
Choose an option to check your answer.
A.
When the variable is numeric
B.
When the sample is complete
C.
When the variable is strongly skewed or contains influential outliers
D.
When the mean is easy to compute
Show Answer
Correct Answer: C. When the variable is strongly skewed or contains influential outliers
Explanation:
The mean can be unrepresentative for skewed data.
Choose an option to check your answer.
A.
Missing at random
B.
Missing completely at random
C.
Missing not at random
D.
Deterministic missingness
Show Answer
Correct Answer: B. Missing completely at random
Explanation:
Under MCAR, missing records form an unbiased subset with respect to the variables.
