MCQs

Disclaimer:

Educational Use Only:

The website links provided are to be used strictly for educational and knowledge-gaining purposes only.

No Misuse:

Do not misuse the links or the information contained therein for any commercial, illegal, or unethical activities.

Respect for Websites and Businesses:

Respect the intellectual property and business interests of the websites and their owners. Do not attempt to alter, copy, or redistribute content without proper authorization.

Acknowledge Value of Knowledge:

Recognize and value the knowledge and resources shared by these websites. Use the information responsibly and ethically.

No Endorsement:

The provision of these links does not imply endorsement of the linked sites, their content, or their products and services. Users are responsible for evaluating the suitability and accuracy of the information.

User Responsibility:

By accessing these links, you agree to use the information at your own risk and to comply with all applicable laws and website terms of use.

Overall MCQs for Reference

1. Which of the following best describes imputation in machine learning?
   a. Dropping missing values
   b. Filling missing values with estimated ones
   c. Normalizing data
   d. Creating a new dataset
   Answer: b. Filling missing values with estimated onessimplilearn​
2. What is the primary goal of imputation?
   a. Remove duplicates
   b. Prevent data leakage
   c. Replace missing data with plausible values
   d. Encode categorical features
   Answer: c. Replace missing data with plausible valuesdataaspirant+1​
3. Which imputation method replaces missing numerical values with the arithmetic average of observed values?
   a. Mode imputation
   b. Median imputation
   c. Mean imputation
   d. Regression imputation
   Answer: c. Mean imputationmitsde+1​
4. Which imputation technique is best suited for categorical variables?
   a. Mean imputation
   b. Median imputation
   c. Mode imputation
   d. KNN imputation
   Answer: c. Mode imputationdataaspirant+1​
5. What type of data does median imputation work best on?
   a. Skewed numerical data
   b. Nominal data
   c. Text data
   d. Ordinal categorical data
   Answer: a. Skewed numerical datamitsde+1​
6. In K-Nearest Neighbors (KNN) imputation, missing values are estimated based on:
   a. Random sampling
   b. Nearest data points in feature space
   c. Regression models
   d. Most frequent categories
   Answer: b. Nearest data points in feature spacedataaspirant+1​
7. What is hot deck imputation?
   a. Using mean from another dataset
   b. Replacing missing values with observed values from similar records
   c. Filling missing values with zero
   d. Using random numbers
   Answer: b. Replacing missing values with observed values from similar recordsmospi+1​
8. Which method involves selection of plausible values for missing data based on the distribution of observed data?
   a. Arbitrary value imputation
   b. Multiple imputation
   c. Mean substitution
   d. One-hot encoding
   Answer: b. Multiple imputationstudy.sagepub+1​
9. Which algorithm is commonly used for model-based imputation?
   a. Logistic regression
   b. Linear regression
   c. Support vector machine
   d. Naive Bayes
   Answer: b. Linear regressiongeeksforgeeks+1​
10. What is the key characteristic of multivariate imputation?
    a. Uses only one feature
    b. Considers correlations among all features
    c. Requires only numerical data
    d. Is identical to mean imputation
    Answer: b. Considers correlations among all featuresscikit-learn+1​
11. Which library provides a range of imputation methods including mean, median, and KNN imputation?
    a. TensorFlow
    b. Scikit-learn
    c. Matplotlib
    d. OpenCV
    Answer: b. Scikit-learnscikit-learn+1​
12. Which imputation method is specifically designed for time-series data, using previous or next values to fill gaps?
    a. Hot deck
    b. Lag imputation
    c. Forward or backward fill
    d. KNN imputation
    Answer: c. Forward or backward fillsimplilearn​
13. In which situation is deleting missing data (listwise deletion) preferable over imputation?
    a. Large proportion of missing data
    b. Random missingness and small dataset
    c. Small proportion of missing data and missingness is completely at random
    d. Data is highly skewed
    Answer: c. Small proportion of missing data and missingness is completely at randomscirp​
14. The EM (Expectation-Maximization) algorithm is typically used for:
    a. Encoding categorical variables
    b. Modeling missing data via iterative parameter estimation
    c. Handling duplicate records
    d. Regularization
    Answer: b. Modeling missing data via iterative parameter estimationaimcqs+1​
15. Which method creates multiple plausible imputations for the same missing value?
    a. Hot deck
    b. MICE (Multiple Imputation by Chained Equations)
    c. Zero imputation
    d. Median imputation
    Answer: b. MICE (Multiple Imputation by Chained Equations)mospi+1​
16. Mean imputation can introduce which bias in the dataset?
    a. Underfitting
    b. Overfitting
    c. Reduced variance
    d. Missing not at random
    Answer: c. Reduced variancescirp​
17. Which imputation technique requires selecting similar records from another, external (historical) dataset?
    a. Hot deck
    b. Cold deck
    c. KNN
    d. Mode imputation
    Answer: b. Cold deckdataaspirant​
18. In KNN imputation, what parameter significantly affects the result?
    a. k (number of neighbors)
    b. alpha value
    c. Distance metric
    d. Both k and distance metric
    Answer: d. Both k and distance metricsciencedirect+1​
19. Imputation should NOT be performed when:
    a. Data is missing at random
    b. A large proportion of the feature is missing
    c. Data is missing not at random (MNAR) with unknown mechanism
    d. The dataset is large
    Answer: c. Data is missing not at random (MNAR) with unknown mechanismscirp​
20. What is the major drawback of single imputation techniques?
    a. Computational complexity
    b. Doesn’t account for uncertainty in missing values
    c. Inability to process categorical data
    d. High variance
    Answer: b. Doesn’t account for uncertainty in missing valuesscirp+1​
21. Which method is best applied for missing categorical values in small datasets?
    a. Mode imputation
    b. Mean imputation
    c. Multivariate imputation
    d. One-hot encoding
    Answer: a. Mode imputationmitsde+1​
22. What should be ensured about missingness before using imputation?
    a. Data is missing at random
    b. No outliers exist
    c. Dataset is balanced
    d. All features are numerical
    Answer: a. Data is missing at randomscirp​
23. Which one is an example of multivariate imputation?
    a. Mode imputation
    b. KNN imputation
    c. Listwise deletion
    d. Median imputation
    Answer: b. KNN imputationsciencedirect+1​
24. What is one limitation of KNN imputation?
    a. Increases dataset variance
    b. Only for categorical data
    c. Computationally expensive with large datasets
    d. Can’t handle missing rows
    Answer: c. Computationally expensive with large datasetssciencedirect+1​
25. Which imputation technique is recommended when the proportion of missing values is very small?
    a. Simple imputation (mean/median/mode)
    b. Multiple imputation
    c. Regression imputation
    d. KNN imputation
    Answer: a. Simple imputation (mean/median/mode)geeksforgeeks+1​
26. To impute missing data using regression, what is required?
    a. Categorical variables
    b. A linear relationship between variables
    c. Random forests
    d. Imputation is not possible
    Answer: b. A linear relationship between variablesgeeksforgeeks+1​
27. Which scenario best fits the use of multiple imputation?
    a. Data missing completely at random with large dataset
    b. Data missing not at random without auxiliary variables
    c. Small dataset with uniform missingness
    d. All continuous features with no correlation
    Answer: a. Data missing completely at random with large datasetstudy.sagepub+1​
28. Which python library provides IterativeImputer for multivariate imputation?
    a. PyTorch
    b. Scikit-learn
    c. Pandas
    d. Seaborn
    Answer: b. Scikit-learnscikit-learn​
29. What is a direct result of improper imputation?
    a. Better model accuracy
    b. Reduced dataset size
    c. Biased model estimates
    d. Decreased training time
    Answer: c. Biased model estimatesscirp​
30. Which statistic is commonly used in mode imputation?
    a. Mean value
    b. Median value
    c. Most frequent value
    d. Least frequent value
    Answer: c. Most frequent valuedataaspirant​
31. The process of replacing missing values with interpolated values between adjacent data points is called:
    a. Regression imputation
    b. Linear interpolation
    c. Forward fill
    d. Hot deck
    Answer: b. Linear interpolationsimplilearn​
32. What defines missing completely at random (MCAR)?
    a. Missingness depends on observed data
    b. Missingness is unrelated to any variable in the dataset
    c. Missingness depends on unobserved data
    d. Missingness depends on feature correlation
    Answer: b. Missingness is unrelated to any variable in the datasetscirp​
33. What type of missing data is handled poorly by standard imputation techniques?
    a. MCAR
    b. MAR
    c. MNAR (Missing Not At Random)
    d. Balanced missing data
    Answer: c. MNAR (Missing Not At Random)scirp​
34. What is the primary risk of imputing missing values in a test dataset?
    a. Data leakage
    b. Loss of model interpretability
    c. Increased missingness
    d. Data imbalance
    Answer: a. Data leakagemitsde​
35. What outcome can be expected if imputation is not performed for missing data before model training?
    a. Decreased computational resources
    b. Errors or lower model accuracy
    c. Increased feature selection capability
    d. Improved data quality
    Answer: b. Errors or lower model accuracymitsde+1​
36. Which imputation method may artificially lower dataset variance?
    a. Multiple imputation
    b. Mean imputation
    c. KNN imputation
    d. Regression imputation
    Answer: b. Mean imputationscirp​
37. Why should imputation be performed after data splitting (into train and test)?
    a. To avoid target leakage
    b. To maximize dataset size
    c. To minimize overfitting
    d. Both a and c
    Answer: d. Both a and cmitsde+1​
38. Which imputation method can use clustering techniques to group similar records for more accurate estimation?
    a. Hot deck
    b. Regression
    c. KNN
    d. MICE
    Answer: a. Hot deckmospi​
39. Which method is preferred for missing data in highly skewed distributions?
    a. Mean imputation
    b. Mode imputation
    c. Median imputation
    d. Zero imputation
    Answer: c. Median imputationdataaspirant+1​
40. Which is a limitation of mode imputation?
    a. Only works for continuous data
    b. May introduce frequency bias in small datasets
    c. Reduces dataset size
    d. Requires external data
    Answer: b. May introduce frequency bias in small datasetsdataaspirant+1​
41. Which type of imputation is designed for continuously updating missing values as more data is collected?
    a. Recursive imputation
    b. Online imputation
    c. Real-time updating
    d. Batch processing
    Answer: b. Online imputationscirp​
42. The ability to perform imputation depends mostly on:
    a. Size of the dataset
    b. Missing data mechanism and model assumptions
    c. Algorithm speed
    d. Feature engineering techniques
    Answer: b. Missing data mechanism and model assumptionsscirp​
43. Which imputation method can handle both categorical and continuous data efficiently?
    a. KNN imputation
    b. Mean imputation
    c. Median imputation
    d. Mode imputation
    Answer: a. KNN imputationmitsde+1​
44. What is the “cold deck” imputation approach?
    a. Uses values from unrelated sources
    b. Uses historical or external datasets for imputation
    c. Interpolates between values in time series
    d. Fills with minimum value
    Answer: b. Uses historical or external datasets for imputationdataaspirant​
45. The presence of which value makes imputation necessary in a dataset?
    a. Zeroes
    b. Null/NaN/missing
    c. Large outliers
    d. Duplicate values
    Answer: b. Null/NaN/missingdataaspirant​
46. What is the impact of ignoring missing values in training data?
    a. No impact
    b. May create model bias or loss of information
    c. Increases accuracy
    d. Removes redundancy
    Answer: b. May create model bias or loss of informationscirp​
47. Which type of imputation is performed where each missing value is imputed only once?
    a. Simple/single imputation
    b. Multiple imputation
    c. Regression imputation
    d. Hot deck imputation
    Answer: a. Simple/single imputationdataaspirant+1​
48. What step usually follows after mean imputation?
    a. Hyperparameter tuning
    b. Feature encoding
    c. Model training or further preprocessing
    d. Tokenization
    Answer: c. Model training or further preprocessingdataaspirant​
49. In machine learning workflows, when should feature imputation be performed?
    a. Before scaling/normalization
    b. After feature selection
    c. After splitting into train and test sets
    d. Whenever convenient
    Answer: a. Before scaling/normalizationmitsde​
50. What does MICE stand for in data imputation?
    a. Mean Imputation and Categorical Encoding
    b. Most Instructive Categorical Estimator
    c. Multiple Imputation by Chained Equations
    d. Multivariate Imputation using Collaborative Estimation
    Answer: c. Multiple Imputation by Chained Equationsmospi+1​

1. https://www.simplilearn.com/data-imputation-article
2. https://dataaspirant.com/data-imputation-techniques/
3. https://blog.mitsde.com/data-imputation-techniques-handling-missing-data-in-machine-learning/
4. https://mospi.gov.in/sites/default/files/publication_reports/A%20Comparison%20of%20Multiple%20Imputation%20Methods%20for%20Data%20with%20missing%20values.pdf
5. https://study.sagepub.com/mehmetogluandjakobsen/student-resources/chapter-13/multiple-choice-quizzes
6. https://www.geeksforgeeks.org/machine-learning/data-imputation-techniques-in-ml/
7. https://scikit-learn.org/stable/modules/impute.html
8. https://www.scirp.org/journal/paperinformation?paperid=137286
9. https://aimcqs.com/unsupervised-learning
10. https://www.sciencedirect.com/science/article/pii/S2352914823002289
11. https://www.sanfoundry.com/1000-machine-learning-questions-answers/
12. https://www.gyansetu.in/blog/machine-learning-mcq-interview-preparation/
13. https://www.igmguru.com/blog/machine-learning-mcqs
14. https://www.geeksforgeeks.org/quizzes/machine-learning-quiz-questions-and-answers/
15. https://diptisurvase.wordpress.com/wp-content/uploads/2022/09/ml-mcqs.pdf
16. https://www.scribd.com/document/520920499/DATA-SCIENCE-AND-MACHINE-LEARNING-MCQ
17. https://www.theanalysisfactor.com/answers-to-the-missing-data-quiz/
18. https://www.aionlinecourse.com/ai-quiz-questions/machine-learning/data-pre-processing
19. https://www.kaggle.com/code/residentmario/simple-techniques-for-missing-data-imputation
20. https://www.kaggle.com/code/azminetoushikwasi/all-imputation-techniques-with-pros-and-cons