Data Mining BSIT 7 Semester/Term PU — University of the Punjab 2025

Question No. 4

Support your Answer with graphical analysis.

CLO 3

10 Marks

a) Why SVM is called maximum margin classifier. Also, justify, why they are critical in determining. the

decision boundary. Draw a 2-D graph showing:

• Two linearly separable classes

• The optimal hyperplane

• Support vectors

• Maximum margin

Explain how SVM selects the optimal hyperplane using diagram.

b)

A university is developing an academic advisory system to classify students into different academic

support categories based on their performance and behavior. The following rule set is used for

classification:

R1: (Attendance = High) ^ (CGPA ≥3.5) → Excellent Student

• R2: (Attendance = Medium) ^ (CGPA ≥ 3.0) → Good Student

• R3: (Attendance = Low) ^ (CGPA < 2.5) → At-Risk Student

• R4: (Assignment Submission = Late) ^ (Attendance = Low) → At-Risk Student

S2

S3

S4

Student

Attendance

High

Medium

Low

Low

CGPA | Assignment Submission

_Applicable Rule(s)

3.7

3.2

2.4

3.1

On-Time

?

Late

On-Time

?

Late

Predicted Class

?

•

Apply the appropriate rule(s) to each student.

Assign the predicted class.

Mention the rule number used for classification.

If more than one rule applies or no rule applies, identify the problem and suggest a solution.

Question No. 5

CLO 3

Recall your semester Project and answer the following questions.

10 Marks

a) Draw a diagram showing steps of your project, showing each step including preprocessing, features,

interpretability analysis (if carried out). Then just write which algorithm shown best result on which

data and what was its value?

b) Write sample piece of code where data is loaded and then one Data mining algorithm is called,

data split is carried out and confusion matrix to be shown. (Hint: use fit and predict functions)

© GOOD LUCK ®

ustadni.com

ustadni.com