Data Mining Algorithms - MCQ Practice

Questions
1. 1. What is the main purpose of association rules in data mining?

 A. To create database indexes

 B. To detect noise in data
 C. To identify relationships between data items
 D. To sort data alphabetically

✅ Answer: C

2. 2. Which of the following is NOT a typical use case of association rules?

 A. Diagnosing medical conditions

 B. Optimizing website interfaces
 C. Designing better sales strategies
 D. Encrypting data

✅ Answer: D

3. 3. In association rule mining, what does the "antecedent" refer to?

 A. The item found with the consequent

 B. The item that appears first in a transaction
 C. The item that leads to the consequent (the “if” part)
 D. The frequency of itemset occurrence

✅ Answer: C

4. 4. What is the measure called that shows how frequently items appear together in the
dataset?

 A. Confidence
 B. Support
 C. Lift
 D. Strength

✅ Answer: B

5. 5. A rule appears frequently in the dataset but rarely holds true when applied. What
does this indicate?
 A. High support, high confidence
 B. Low support, low confidence
 C. High support, low confidence
 D. Low support, high confidence

✅ Answer: C

6. 6. What does a lift value of 1 indicate in association rule mining?

 A. Negative correlation
 B. Positive correlation
 C. No correlation
 D. High support

✅ Answer: C

7. 7. Which algorithm generates candidate itemsets by joining large itemsets from the
previous pass with themselves?

 A. AIS
 B. SETM
 C. Apriori
 D. K-means

✅ Answer: C

8. 8. Which algorithm saves the transaction ID with each candidate itemset?

 A. AIS
 B. Apriori
 C. Decision Tree
 D. SETM

✅ Answer: D

9. 9. Why is the Apriori algorithm more efficient than AIS and SETM?

 A. It uses fewer variables

 B. It only explores itemsets that meet minimum support
 C. It doesn't scan the database
 D. It ignores confidence values

✅ Answer: B

10. 10. In which business application is market basket analysis commonly used?
 A. Employee training
 B. Product recommendation
 C. Supply chain management
 D. Retail sales

✅ Answer: D

11. 11. In the diaper and beer example, what percentage of transactions included both
items?

 A. 2.75%
 B. 2%
 C. 1.75%
 D. 0.5%

✅ Answer: C

12. 12. What does a high confidence and low support in an association rule imply?

 A. The rule is common but often incorrect

 B. The rule is rare but often correct
 C. The rule is common and always correct
 D. The rule is never useful

✅ Answer: B

13. 13. Which of the following is the correct form of an association rule?

 A. if [condition] then [result]

 B. for [every] do [action]
 C. select * from [database]
 D. input [x] to get [y]

✅ Answer: A

14. 14. What is the first step in building a classification model?

 A. Testing with unlabeled data

 B. Using normalization techniques
 C. Creating a classifier with labeled training data
 D. Drawing a pie chart

✅ Answer: C

15. 15. Which of the following best describes classification?

 A. Predicting numerical outcomes
 B. Removing duplicate data entries
 C. Assigning a category label to new observations
 D. Encrypting sensitive information

✅ Answer: C

16. 16. What is the output of a classification model?

 A. A numerical estimate
 B. A set of decision rules
 C. A category label
 D. A list of all itemsets

✅ Answer: C

17. 17. What is the major difference between classification and prediction?

 A. Classification uses regression; prediction does not

 B. Classification is unsupervised; prediction is supervised
 C. Classification predicts categories; prediction estimates numerical values
 D. Prediction uses only text data

✅ Answer: C

18. 18. Which of the following is an example of a classification algorithm?

 A. Linear Regression
 B. DBSCAN
 C. K-Means
 D. Naive Bayes

✅ Answer: D

19. 19. In data preparation, which method helps scale values into a small specified range?

 A. Generalization
 B. Noise Reduction
 C. Normalization
 D. Aggregation

✅ Answer: C

20. 20. Which algorithm is best suited for predicting continuous values?
 A. Logistic Regression
 B. K-Nearest Neighbors
 C. Linear Regression
 D. Decision Tree Classification

✅ Answer: C

21. 21. Which is *not* a stage in the data classification lifecycle?

 A. Storage
 B. Forecasting
 C. Sharing
 D. Publication

✅ Answer: B

22. 22. Which of the following algorithms is used in both classification and prediction?

 A. Logistic Regression
 B. Apriori
 C. DBSCAN
 D. K-Means

✅ Answer: A

23. 23. What does a decision tree output when used for classification?

 A. A list of association rules

 B. A numerical score
 C. A continuous value
 D. A categorical decision based on branches

✅ Answer: D

24. 24. What technique is used to identify whether two attributes are related?

 A. Clustering
 B. Correlation analysis
 C. Filtering
 D. Regression

✅ Answer: B

25. 25. What is the goal of relevance analysis in data preparation?

 A. Create predictive models
 B. Remove noise
 C. Identify useful attributes
 D. Encode categorical data

✅ Answer: C

26. 26. What fundamental principle does the Naive Bayes algorithm rely on?

 A. Decision boundaries
 B. Baye’s theorem
 C. Distance measures
 D. Gaussian distribution

✅ Answer: B

27. 27. What type of learning method is clustering considered?

 A. Supervised
 B. Reinforcement
 C. Unsupervised
 D. Semi-supervised

✅ Answer: C

28. 28. In a shopping mall, grouping similar items together like t-shirts or vegetables is an
example of:

 A. Supervised learning
 B. Data labeling
 C. Clustering
 D. Classification

✅ Answer: C

29. 29. Which of the following is an application of clustering?

 A. Data encryption
 B. Market segmentation
 C. Linear regression
 D. Data normalization

✅ Answer: B
30. 30. Which clustering method does NOT require the number of clusters to be specified in
advance?

 A. K-Means
 B. Partitioning Clustering
 C. Hierarchical Clustering
 D. Fuzzy Clustering

✅ Answer: C

31. 31. Which clustering method allows data points to belong to more than one cluster?

 A. Hard clustering
 B. K-Means
 C. Agglomerative Hierarchical
 D. Fuzzy Clustering

✅ Answer: D

32. 32. Which algorithm is an example of a density-based clustering model?

 A. DBSCAN
 B. K-Means
 C. Naive Bayes
 D. Decision Tree

✅ Answer: A

33. 33. What is a limitation of density-based clustering methods like DBSCAN?

 A. Requires pre-labeled data

 B. Can't handle outliers
 C. Difficulty with varying densities and high dimensions
 D. Needs class labels

✅ Answer: C

34. 34. What type of model does the Expectation-Maximization (EM) algorithm use in
clustering?

 A. Centroid-based
 B. Rule-based
 C. Distribution-based
 D. Tree-based
 ✅ Answer: C

35. 35. What type of clustering algorithm is K-Means?

 A. Density-based
 B. Distribution-based
 C. Partitioning
 D. Hierarchical

✅ Answer: C

36. 36. Which algorithm performs bottom-up hierarchical clustering?

 A. K-Means
 B. DBSCAN
 C. Agglomerative Hierarchical
 D. Expectation-Maximization

✅ Answer: C

37. 37. Which algorithm avoids the need to define the number of clusters and works by
message passing?

 A. K-Means
 B. Affinity Propagation
 C. DBSCAN
 D. Mean-shift

✅ Answer: B

38. 38. What is the most common use of clustering in biology?

 A. Labeling DNA sequences

 B. Identifying cancerous cells
 C. Sorting email messages
 D. Constructing protein molecules

✅ Answer: B

39. 39. In clustering, what does “soft clustering” mean?

 A. Data is split randomly

 B. Each data point belongs to exactly one group
 C. Each data point may belong to multiple groups
 D. Clusters are based on soft data types
 ✅ Answer: C

40. 40. What does the Mean-Shift algorithm do?

 A. Sorts data into fixed clusters

 B. Finds dense regions and updates centroids
 C. Uses decision trees for classification
 D. Labels data using Naive Bayes

✅ Answer: B