21EC744

Model Question Paper-1/2 with effect from 2021(CBCS Scheme)

USN

7th Semester B.E. Degree Examination

Subject Title : Machine Learning with Python
TIME: 03 Hours Max. Marks: 100

Note: 01. Answer any FIVE full questions, choosing at least ONE question from each
MODULE.

*Bloom’s COs
Module -1 Taxonomy Marks
Level
Q.01 a Discuss types of Machine Learning with examples. L1 1 10M
b Illustrate the adaptive linear neurons(Adaline) in Machine Learning L2 1 10M
system
OR
Q.02 a Describe the process in building machine learning systems. L1 1 10M

b Illustrate the perceptron learning Rule in Machine Learning system L2 1 10M

Module-2
Q. 03 a Describe how to train a perceptron using Scikit Learn L1 2 10M

b Illustrate the one hot encoding on nominal features for a given data L2 2 10M
OR
Q.04 a Describe implementation of Logistic Regression using Scikit Learn L1 2 10M

b Illustrate the encoding class labels on supervised data. L2 2 10M

Module-3
Q. 05 a Apply Principle Component Analysis on unsupervised data for L3 3 10M
dimensionality reduction

b Apply the bag-of-words model to the IMDb movie review dataset L3 3 10M
to convert textual data into numerical features suitable for machine
learning.
OR
Q. 06 a Apply Linear Discriminant Analysis on supervised data for L3 3 10M
dimensionality reduction

b Apply a logistic regression model to the IMDb movie review dataset L3 3 10M
and classify reviews as positive or negative using the bag-of-words
representation.

Module-4
Q. 07 a Describe how to setup on SQL Lite database for data storage. L1 4 10M

b L3 4 10M
Solve for modeling nonlinear relationships in Housing dataset
OR
Q. 08 a Describe turning the movie classifier into web application. L1 4 10M
b Solve for estimating the coefficient of a regression model via scikit L3 4 10M
learn.

@#@
Page 01 of 02
 21EC744
Module-5
Q. 09 a Apply polynomial regression to extend a linear regression model L3 5 10M
into a curve and compare its performance against standard linear
regression using the Housing Dataset
b Apply a simple linear regression model to the Housing Dataset to L3 5 10M
predict housing prices and evaluate its performance using
appropriate metrics.
OR
Q. 10 a Apply k-means clustering algorithm to group objects by similarity L3 5 10M
and analyse the resulting clusters.
b Apply an artificial neural network to model complex functions and L3 5 10M
classify handwritten digits with high accuracy.

*Bloom’s Taxonomy Level: Indicate as L1, L2, L3, L4, etc. It is also desirable to indicate the COs and POs to be
attained by every bit of questions.

@#@
Page 02 of 02
 21EC744
Model Question Paper-2/2 with effect from 2021(CBCS Scheme)
USN

7th Semester B.E. Degree Examination

Subject Title : Machine Learning with Python
TIME: 03 Hours Max. Marks: 100

Note: 01. Answer any FIVE full questions, choosing at least ONE question from each
MODULE.

*Bloom’s COs
Module -1 Taxonomy Marks
Level
Q.01 a Illustrate minimizing cost functions with gradient descent . L2 1 10M
b Describe how to implement a perceptron learning algorithm on Iris L1 1 10M
dataset
OR
Q.02 a Illustrate how to improve gradient descent through feature scaling L2 1 10M

b Describe adaptive linear neurons (Adaline) with minimizing cost L1 1 10M

functions with gradient descent.
Module-2
Q. 03 a Apply pre-processing techniques to clean a given dataset having L3 2 10M
missing values.

b Describe maximum margin classification with support vector machine L1 2 10M

with example.

OR
Q.04 a Apply k-NN algorithm to classify multiclass -unsupervised data. L3 2 10M

b Describe how to handle categorical data with example . L1 2 10M

Module-3
Q. 05 a Apply the bag-of-words model to the IMDb movie review dataset L3 3 10M
to convert textual data into numerical features suitable for machine
learning.
b Explain confusion matrix, precision , recall ,F1 score ,ROC with L1 3 10M
example

OR
Q. 06 a Apply a logistic regression model to the IMDb movie review dataset L3 3 10M
and classify reviews as positive or negative using the bag-of-words
representation.

b Describe the process of k-fold cross-validation . Why it is preferred L1 3 10M

over a single train-test split ?

Module-4
Q. 07 a Describe the key components and workflow of developing a web L1 4 10M
application using flask.

b Apply regularization techniques such as ridge regression ,lasso L3 4 10M

regression to address overfitting in a regression model.

@#@
Page 01 of 02
 21EC744
OR
Q. 08 a Describe deploying the web application to public server. L1 4 10M
b Apply RANSAC algorithm to fit a robust regression model on a L3 4 10M
dataset and discuss its ability to handle outliers and improve model
performance .
Module-5
Q. 09 a Apply polynomial regression to extend a linear regression model L3 5 10M
into a curve and compare its performance against standard linear
regression using the Housing Dataset
b Apply a simple linear regression model to the Housing Dataset to L3 5 10M
predict housing prices and evaluate its performance using
appropriate metrics.
OR
Q. 10 a Apply k-means clustering algorithm to group objects by similarity L3 5 10M
and analyse the resulting clusters.
b Apply with different neural network architectures to improve L3 5 10M
performance on handwritten digit classification tasks.

*Bloom’s Taxonomy Level: Indicate as L1, L2, L3, L4, etc. It is also desirable to indicate the COs and POs to be
attained by every bit of questions.

@#@
Page 02 of 02
 21EC744
Model Question Paper-1 with effect from 2021(CBCS Scheme)
USN

7th Semester B.E. Degree Examination

Machine Learning with Python
TIME: 03 Hours Max. Marks: 100

Note: 01. Answer any FIVE full questions, choosing at least ONE question from each
MODULE.

*Bloom’s COs
Module -1 Taxonomy Marks
Level
Q.01 a Explain the different types of learning. L2 CO1 8
b Discuss the roadmap for building machine learning systems. L2 CO1 7
c Explain the biological neuron with a neat diagram. L2 CO1 5
OR
Q.02 a Explain the different types of learning. L2 CO1 8
b Explain the Adaptive linear neurons and the convergence of learning. L2 CO1 7
c Explain the biological neuron with a neat diagram. L2 CO1 5
Module-2
Q. 03 a Explain the concepts overfitting and underfitting problem in machine CO1
L2 8
learning.
b How does Logistic Regression handle multi-class classification? L2 CO1 7
c What are L1 and L2 regularization, and when would you use them? L2 CO4 5
OR
Q.04 a What is the geometric interpretation of the "margin" in SVMs? L2 CO3 8
b Demonstrate k- Nearest Neighbors with example. L2 CO4 7
c Why are kernels used in SVMs? L2 CO4 5
Module-3
Q. 05 a Discuss principal component analysis method. L2 CO1 8
b Explain the inner workings of linear discriminant analysis. L2 CO1 7
c Interpret how to assess feature importance with random forests? L2 CO1 5
OR
Q. 06 a Explain partitioning a dataset into training and test dataset. L2 CO3 8
b How to deal with missing data and categorical data? L2 CO3 7
c Discuss about bag of words L2 CO4 5
Module-4
Q. 07 a Develop a python code to create a new SQLite database inside the
L3 CO3 10
movieclassifier directory and store two example movie reviews.
b Explain the mathematical formulation of a simple linear regression model
L2 CO3 10
with a neat diagram.
OR
Q. 08 a Describe the algorithmic steps involved in fitting a robust regression
L2 CO2 10
model using RANSAC.
b What metrics are commonly used to evaluate the performance of linear
L2 CO2 10
regression models? Discuss their significance.
Module-5
Q. 09 a What is the k-means clustering algorithm, and how does it group objects
L2 CO2 10
based on similarity?
b What is hierarchical clustering, and demonstrate the working with the
L2 CO2 10
help of an example?

@#@
Page 01 of 02
 21EC744
OR
Q. 10 a Explain the working of multilayer neural network architecture. L2 CO3 8
b Explain the process of backpropagation and how it updates weights
L2 CO4 7
during training.
c What is the MNIST dataset, and why is it widely used in digit
L2 CO4 5
classification tasks?
*Bloom’s Taxonomy Level: Indicate as L1, L2, L3, L4, etc. It is also desirable to indicate the COs and POs to be
attained by every bit of questions.

@#@
Page 02 of 02