Scribd
Upload
188 views3 pages

Al 3451-Machine Learning 2025

This document is an examination question paper for the course 'Machine Learning' at Saranathan College of Engineering, covering various topics in artificial intelligence and data science. It consists of two parts: Part A with 10 short answer questions and Part B with 5 detailed questions, each offering options for students to choose from. The exam is scheduled for three hours and has a maximum score of 100 marks.

Uploaded by

takash2412
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
188 views3 pages

Al 3451-Machine Learning 2025

This document is an examination question paper for the course 'Machine Learning' at Saranathan College of Engineering, covering various topics in artificial intelligence and data science. It consists of two parts: Part A with 10 short answer questions and Part B with 5 detailed questions, each offering options for students to choose from. The exam is scheduled for three hours and has a maximum score of 100 marks.

Uploaded by

takash2412
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Reg. No.

Question Paper Code: 27203

SARANATHAN COLLEGE OF ENGINEERING, TIRUCHIRAPPALLI - 620 012


(AUTONOMOUS)

B.E./B.Tech. END SEMESTER EXAMINATIONS, NOV/DEC 2024

ARTIFICIAL INTELLIGENCE AND DATA SCIENCE

IV Semester

AL3451 - MACHINE LEARNING


(Common to: Computer Science and Engineering (AIML) and Computer Science and Business Systems)
(Regulation 2021)
Time: Three Hours Max. Marks: 100

Answer ALL Questions

PART - A (10 x 2 = 20 Marks)


1. How does overfitting relate to the bias-variance trade-off? CO1 L1

2. What is inductive bias, and how does it affect learning algorithms? CO1 L1

3. Differentiate between single-variable and multiple-variable linear CO2 L2


regression.

4. Define the Naive Bayes classifier. CO2 L1

5. What is ensemble learning, and why is it used in machine learning? CO3 L1

6. Define bagging and discuss its advantages. CO3 L1

7. Why is error backpropagation essential for training multilayer CO4 L1


perceptrons?

8. How does the choice of optimizer impact the training process of a CO4 L1
neural network?

9. Why is it important to use statistical tests like the t-test and CO5 L1
McNemar’s test when comparing classifiers?

10. List the factors that should be considered when interpreting the CO5 L1
results of machine learning experiments.
PART - B (5 x 16 = 80 Marks)
Page 1 of 3
11. (a) Provide a comprehensive overview of linear algebra concepts (16) CO1 L3
relevant to machine learning.
(OR)
11. (b) Describe the hypothesis space in machine learning and Discuss its (16) CO1 L3
implications on the learning process and model selection.

12. (a) Describe how you would build a linear regression model to predict (16) CO2 L4
house prices. Include the following in your answer:
i) The steps involved in data preprocessing (e.g., handling
missing values, feature scaling).
ii) How would you implement the least squares method for
training your model?
iii) Discuss the importance of evaluating your model using metrics
such as R-squared and Mean Absolute Error.
(OR)
12. (b) You are working on a project to classify emails as either "spam" or (16) CO2 L4
"not spam" using features extracted from the email content.
Describe how you would implement a Support Vector Machine
(SVM) for this classification task.

13. (a) You have a dataset of customer transactions and want to segment (16) CO3 L3
customers based on purchasing behavior using K-means clustering.
Explain the steps you would take to implement K-means, including
how to choose the number of clusters and evaluate the clustering
results.
(OR)
13. (b) Discuss different voting mechanisms used in ensemble learning (16) CO3 L3
(e.g., majority voting, weighted voting). How do these mechanisms
affect the performance of the ensemble model?

14. (a) Explain how you would implement stochastic gradient descent to (16) CO4 L3
train a neural network with relevant examples.
(OR)
14. (b) Explain how batch normalization works and discuss its impact on (16) CO4 L3
the training process of neural networks.

15. (a) Compare and contrast K-fold cross-validation and bootstrapping as (16) CO5 L3
Page 2 of 3 27203
methods for assessing model performance and Discuss their
advantages and disadvantages.
(OR)
15. (b) Discuss how you would assess overfitting and underfitting in a (16) CO5 L3
machine learning model. Include the role of cross-validation in this
assessment and how you would adjust your model or experiment
design based on your findings.
*****

Page 3 of 3 27203

You might also like