Data Mining - University of Sargodha Terminal Examinations Fall 2025 Program BS(IT) 7th (Reg & SS1) Part 1

University of Sargodha, Sargodha

Terminal Examinations Fall 2025

Program/Class:_BS(IT) 7th (Reg & SS1)

Subject: Data Mining

Student Name:

Subject: Data Mining_- 50 Marks

Instructor:

Ms. Anam Naz.

Roll#:

Note: Attempt all questions, attempt subparts of a question in a sequence, each of equal marks i.e.,

10. Please

do not write anything on the Question Paper except Name and Roll Number.

10 Marks

Question No. 1

CLO 1

Provide short logical answers.

a)

Between Holdout and Cross validation, which is better and why? Share Pros and cons of both (

approaches for data split.

b) Differentiate between Classification and Regression with the help of example each?

c) Difference between Model and Data, which is more important and why?

d) Why kNN is called Lazy Classifier?

e)

What is reason to carry out anonymization of data?

CLO 2

10 Marks

Question No. 2

Compute the following as required.

A nutritionist wants to analyze the similarity among 4 food supplements based on two nutritional attributes in

order to form two clusters using the K-Means clustering algorithm. Assume

Supplement B and Supplement D are selected as the initial centroids.

Supplement

a) Compute the Euclidean distance of each supplement from the initial

centroids.

b) Assign each supplement to the nearest cluster after Iteration

1.

c) Recalculate the new centroids.

Protein

(g)

12

8

15

10

Energ

(kcal)

180

200

150

160

d) Repeat the process for Iteration 2 and show the final clusters.

CLO 1

10 Marks

Question No. 3

Do Background Analysis.

a) The rapid growth of data generated through transactions, interactions, and observations has led to the

emergence of Big Data-driven data mining techniques. These techniques play a crucial role in

enabling modern emerging technologies. Explain the role of data mining in the following emerging

technologies:

Genetic Engineering

•

Augmented Reality (AR) and Virtual Reality (VR)

•

Internet of Things (loT)

Explain how data mining helps in pattern discovery, prediction, and decision-making in each domain.

b) Using the concept of the 3V's of Big Data supports, explain a data flow that shows:

Growth of data size from Megabytes (MB) → → Petabytes (PB)

• Increase in data complexity and diversity by analyzing transition from traditional structured data

(e.g., ERP, CRM) to web, sensor, and social media data

ustadni.com