Scribd
Upload
13 views7 pages

Data Mining MCQs - Unit 4

The document contains multiple-choice questions (MCQs) related to Data Mining, specifically focusing on clustering techniques and algorithms such as k-means and DBSCAN. It includes questions about definitions, methods, and characteristics of clustering, as well as fill-in-the-blank questions for further assessment. This resource serves as a study guide for students in the field of Data Mining.

Uploaded by

e3hodcse
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
13 views7 pages

Data Mining MCQs - Unit 4

The document contains multiple-choice questions (MCQs) related to Data Mining, specifically focusing on clustering techniques and algorithms such as k-means and DBSCAN. It includes questions about definitions, methods, and characteristics of clustering, as well as fill-in-the-blank questions for further assessment. This resource serves as a study guide for students in the field of Data Mining.

Uploaded by

e3hodcse
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance

 Menu

Objective Type Questions & Answers

Data Mining MCQs - Unit-4

1. k-means is one of the simplest _______________ learning algorithms.

A . Supervised

B . Unsupervised

C . Both

D . None

Answer

2. Clustering is a_________________________________

A . Group of similar objects that differ significantly from other objects

B . Operations on a data to transform data in order to prepare it for a data mining algorithm

C . Symbolic representation of facts from which information can potentially be extracted

D . None

Answer

3. ____________________ method is used for partitioning the clusters based on the medoid objects.

A . k-means

B . k-medoid

C . Divisive

D . DBSCAN

Answer

4. ______________ hierarchical clustering algorithm initiates with clusters and integrates the two closest clusters until only one cluster
remains.

A . DBSCAN

B . Agglomerative

C . k-means

D . Divisive

Answer

5. In ___________________ algorithm, each cluster is represented by the mean value of the objects in the cluster.

[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 1/7


18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance

A . k-means

B . k-medoid

C . DBSCAN

D . Hierarchal

Answer

6. Which of the following is not a requirement of clustering?

A . Scaling of data

B . Large dimensionality

C . Divisive clustering

D . Arbitrary shaped cluster detection

Answer

7. _________method works by grouping data objects into a tree of clusters

A . Partition based clustering

-65% Kortex 2600W Red -8% Tiger Lorex Steel -84% Yoton UC500
Portable High Toe PU Sole Black 400lm Yellow
Pressure Car Safety Shoes"|" Portable Mini
Washer with All… Size: 8 Home Theater L…

Rs4,449 Rs1,139 Rs1,599


Go ! Go ! Go !

-76% Cybernautt C2K -58% ArmaDuro AD1003 -35% Allen Cooper AC-
2MP HD Magnetic Suede Leather 1156 Antistatic
Smart WiFi Steel Toe Tan Steel Toe Grey &
Camera with Nig… Safety Shoes"|"… Black Safety Sh…

Rs599 Rs668 Rs1,691


Go ! Go ! Go !

B . Grid based clustering

C . Density based clustering

D . Hierarchical based clustering

Answer

8. A distance between different cluster is measured through___________

A . Linkage metric

B . Relation metric

C . Weight metric

D . None of the above

[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 2/7


18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance

Answer

9. ____________ is a technique of combining a group of physical objects into classes of homogeneous objects.

A . Clustering

B . Grapping

C . Bisecting

D . Optimizing

Answer

10.) K-medoids clustering methods uses one representative________ per cluster to represent the clusters.

A . Mean

B . Median

C . Object

D . Space

Answer

11.) The occurrences of_________ is inevitable when Euclidean distance criterion is employed.

A . Outlier

B . Empty cluster

C . Centroids

D . DBSCAN

Answer

[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 3/7


18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance
Fill in the Blanks

12. Cluster contains a set of_______ data objects.

Answer

13. Huge datasets are partitioned into groups based on___________.

Answer

14. _____________________________methods can find arbitrary shaped clusters.

Answer

15. DBSCAN stands for_____________________________________________________.

Answer

16. Outliers are different from noisy data (True/False)

Answer

17. _______________________________ is a "bottom-up" approach: each observation starts in its own cluster, and pairs of clusters
are merged as one moves up the hie

Answer

18. ________________________ distance measure is used to find similarity among data points in k-means algorithm.

Answer

19. _________________ algorithm is efficient for clustering even if the dataset contains outliers.

Answer

[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 4/7


18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance

20. _______________________________ is a "top-down" approach in hierarchical clustering m

Answer

Relevant Materials :

☞ Data Mining MCQs - Unit-1 - [ DM ]

☞ Data Mining MCQs - Unit-2 - [ DM ]


☞ Data Mining MCQs - Unit-3 - [ DM ]
☞ Data Mining MCQs - Unit-4 - [ DM ]
☞ Data Mining MCQs - Unit-5 - [ DM ]

Similar Materials :

☞ PPS MCQs - Unit-1 - [ PPS ]


☞ PPS MCQs - Unit-2 - [ PPS ]
☞ PPS MCQs - Unit-3 - [ PPS ]
☞ PPS MCQs - Unit-4 - [ PPS ]
☞ PPS MCQs - Unit-5 - [ PPS ]
☞ Object Oriented Programming through Java MCQs - Unit-1 - [ OOP_JAVA ]
☞ Object Oriented Programming through Java MCQs - Unit-2 - [ OOP_JAVA ]
☞ Object Oriented Programming through Java MCQs - Unit-3 - [ OOP_JAVA ]

[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 5/7


18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance

☞ Object Oriented Programming through Java MCQs - Unit-4 - [ OOP_JAVA ]


☞ Object Oriented Programming through Java MCQs - Unit-5 - [ OOP_JAVA ]
☞ Design and Analysis of Algorithms MCQs - Unit-1 - [ DAA ]
☞ Design and Analysis of Algorithms MCQs - Unit-2 - [ DAA ]
☞ Design and Analysis of Algorithms MCQs - Unit-3 - [ DAA ]
☞ Design and Analysis of Algorithms MCQs - Unit-4 - [ DAA ]
☞ Design and Analysis of Algorithms MCQs - Unit-5 - [ DAA ]
☞ Software Engineering MCQs - Unit-1 - [ SE ]
☞ Software Engineering MCQs - Unit-2 - [ SE ]
☞ Software Engineering MCQs - Unit-3 - [ SE ]
☞ Software Engineering MCQs - Unit-4 - [ SE ]
☞ Software Engineering MCQs - Unit-5 - [ SE ]
☞ Computer Organization and Architecture (COA) Objective Question Bank-Unit-1 - [ COA ]
☞ Computer Organization and Architecture (COA) Objective Question Bank-Unit-2 - [ COA ]
☞ Computer Organization and Architecture (COA) Objective Question Bank-Unit-3 - [ COA ]
☞ Computer Organization and Architecture (COA) Objective Question Bank-Unit-4 - [ COA ]
☞ Computer Organization and Architecture (COA) Objective Question Bank-Unit-5 - [ COA ]
☞ Data Structures Objective Type Question Bank-Unit-1 - [ DS ]
☞ Data Structures Objective Type Question Bank-Unit-2 - [ DS ]
☞ Data Structures Objective Type Question Bank-Unit-3 - [ DS ]
☞ Data Structures Objective Type Question Bank-Unit-4 - [ DS ]
☞ Data Structures Objective Type Question Bank-Unit-5 - [ DS ]
☞ Database Management System Objective Type Question Bank-Unit-1 - [ DBMS ]
☞ Database Management System Objective Type Question Bank-Unit-2 - [ DBMS ]
☞ Database Management System Objective Type Question Bank-Unit-3 - [ DBMS ]
☞ Database Management System Objective Type Question Bank-Unit-4 - [ DBMS ]
☞ Database Management System Objective Type Question Bank-Unit-5 - [ DBMS ]
☞ Cyber Forensics Objective Type Question Bank-Part-2 - [ Cyber Forensics ]
☞ Cyber Forensics Objective Type Question Bank-Part-1 - [ Cyber Forensics ]
☞ Java Programming Objective Type Question Bank - [ Java Programming ]
☞ Java Programming Objective Type Questions-Part-1 - [ Java Programming ]
[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 6/7
18/01/2025, 11:12 Data Mining MCQs - Unit-4 - DM | Study Glance

☞ Java Programming Objective Type Questions-Part-2 - [ Java Programming ]


☞ Java Programming Objective Type Questions-Part-3 - [ Java Programming ]
☞ Java Programming Objective Type Questions-Part-4 - [ Java Programming ]
☞ Java Programming Objective Type Questions-Part-5 - [ Java Programming ]
☞ Neural Networks and Deep Learning-Unit-1 Objective Type Questions - [ NNDL ]
☞ Neural Networks and Deep Learning-Unit-2 Objective Type Questions - [ NNDL ]
☞ Neural Networks and Deep Learning-Unit-3 Objective Type Questions - [ NNDL ]
☞ Neural Networks and Deep Learning-Unit-4 Objective Type Questions - [ NNDL ]
☞ Operating Systems -Unit-1 Objective Type Questions - [ Operating Systems ]
☞ Operating Systems -Unit-2 Objective Type Questions - [ Operating Systems ]
☞ Operating Systems -Unit-3 Objective Type Questions - [ Operating Systems ]

This site uses Google AdSense ad intent links. AdSense automatically generates these links and they may help creators earn money.

ABOUT
Study Glance provides Tutorials , Power point Presentations(ppts), Lecture Notes, Important & previously asked questions, Objective
Type questions, Laboratory programs and we provide Syllabus of various subjects.

CATEGORIES
Tutorials Questions
PPTs Lab Programs
Lecture Notes Syllabus

Copyright © 2020 All Rights Reserved by StudyGlance.

   

[Link] Mining MCQs - Unit-4&subject=DM&furl=data/btech/btechobjquestio… 7/7

Common questions

Powered by AI

DBSCAN is preferred over k-means clustering when the dataset includes noise and outliers, or when the shapes of clusters are non-spherical. DBSCAN's main advantages are its ability to detect clusters of arbitrary shapes and its robustness against noise, as it does not require specifying the number of clusters beforehand, unlike k-means which assumes spherical clusters and a fixed number of clusters .

Selecting an appropriate distance measure is crucial as it determines the shape of clusters that can be identified and affects the clustering results. For instance, Euclidean distance is typically used in algorithms like k-means but may be inappropriate for clusters of arbitrary shapes or when outliers are present because it assumes spherical clusters. Alternatively, Manhattan or cosine distance measures can be used when data is sparse or when angles and orientations are more significant .

Clustering high-dimensional data sets poses several challenges, such as the curse of dimensionality, increased computational complexity, and reduced cluster separation. Methods to address these challenges include dimensionality reduction techniques like Principal Component Analysis (PCA) and feature selection, which help in reducing the dimensionality while preserving clustering-relevant structures .

Agglomerative hierarchical clustering begins with each observation in its own cluster and repeatedly merges the closest pair of clusters until only one cluster remains . Conversely, divisive clustering is a "top-down" approach that starts with all observations in a single cluster and successively splits them until each observation is in its own cluster .

Cluster scalability impacts algorithm choice, as some algorithms, such as k-means, struggle with very large datasets due to increased computation time. Strategies to manage scalability include parallel processing, reducing dimensionality, using incremental clustering methods, or applying more scalable algorithms like DBSCAN, which can handle larger datasets by efficiently processing only dense areas using less memory and computational power .

Outliers can significantly distort the results of clustering algorithms, leading to inaccurate cluster representations. Proper handling of outliers is crucial to improving cluster accuracy. Methods to manage outliers include density-based clustering algorithms like DBSCAN, which can identify and separate dense areas of data from sparse outliers, and statistical techniques that detect and remove abnormalities based on deviation from the norm .

A linkage criterion in hierarchical clustering defines the distance measure between clusters to determine which clusters to merge at each step. The choice of linkage criterion, such as single, complete, or average linkage, influences cluster shapes and sizes, potentially affecting the hierarchy's outcome, as different criteria can result in different dendrogram representations and cluster assignments .

Supervised learning involves models trained on labeled data where the output is known and used to make predictions, whereas unsupervised learning, including clustering, deals with unlabeled data to identify inherent structures or patterns. Clustering, as an unsupervised learning method, is useful for exploratory data analysis where class labels are not predetermined .

K-means clustering represents each cluster by the mean value of the objects in the cluster, which works well for spherical clusters but is sensitive to outliers. On the other hand, k-medoids uses one representative object (medoid) per cluster to represent the clusters, which makes it more robust to noise and outliers as it minimizes a sum of pairwise dissimilarities instead of variance .

Preprocessing plays a critical role in clustering by preparing the data for more effective analysis. Common preprocessing techniques include normalization and scaling, which help in adjusting the data into a standard range, and noise reduction through methods like outlier detection, which can reduce the impact of anomalies on clustering results .

You might also like