Scribd
Upload
4 views3 pages

Machine Learning Assignment 2024-25

This document is an assignment for B.Sc(H) Computer Science students at Deen Dayal Upadhyaya College, focusing on various machine learning concepts. It includes questions on clustering methods, classification metrics, Naïve Bayes predictions, Support Vector Machines, hierarchical clustering, bias-variance tradeoff, neural networks, and the curse of dimensionality. The assignment is due on April 17, 2025, and each question carries 5 marks.

Uploaded by

kitobed370
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
4 views3 pages

Machine Learning Assignment 2024-25

This document is an assignment for B.Sc(H) Computer Science students at Deen Dayal Upadhyaya College, focusing on various machine learning concepts. It includes questions on clustering methods, classification metrics, Naïve Bayes predictions, Support Vector Machines, hierarchical clustering, bias-variance tradeoff, neural networks, and the curse of dimensionality. The assignment is due on April 17, 2025, and each question carries 5 marks.

Uploaded by

kitobed370
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

Deen Dayal Upadhyaya College

[Link](H) Computer Science VI Sem (2024-25)


Assignment: Machine Learning

DeadLine : 17/04/2025 1:00 pm


Each question carry 5 marks

Q1 What is the difference between K-Means and Hierarchical Clustering? In what


situations would you prefer one over the other?

Q2 Define the following terms in the context of classification: Precision,


Recall/sensitivity, F1 Score. What do ROC and AUC represent in a classification
task? Why are they useful?

Q3 Using Naïve Bayes classification rule for the following training data, predict whether
an old student having medium height will have the capability to swim or not?​
​ ​ ​ ​ ​ ​
Id Age Height Student Can swim
1. Young High No No
2. Young High No No
3. Middle High No Yes
4. Old Medium No Yes
5. Old Low Yes Yes
6. Old Low Yes No
7. Middle Low Yes Yes
8. Young Medium No No
9. Young Low Yes Yes
10. Old Medium Yes Yes
11. Young Medium Yes Yes
12. Middle Medium No Yes
13. Middle High Yes Yes
14. Old Medium No No

Q4 What is the role of the kernel in Support Vector Machines (SVM)? Briefly explain
linear and RBF kernels.

Q5 Differentiate between hard margin and soft margin SVM. When would you use each?
What is the role of the regularization parameter C in SVM?

Q6 You are given the following 5 data points and their pairwise distances:
A B C D E

A 0 2 6 10 9

B 2 0 5 9 8

C 6 5 0 4 5

D 10 9 4 0 3

E 9 8 5 3 0

1.​ Perform Agglomerative Hierarchical Clustering using Single Linkage


(minimum distance).​
2.​ Show the step-by-step merging of clusters and corresponding distances.​
3.​ Draw the dendrogram to illustrate the process.

Q7 What is the bias-variance tradeoff? How is it related to overfitting and underfitting?


What symptoms in your training and validation accuracy would indicate overfitting ?
List at least 3 techniques to prevent or reduce overfitting in machine learning models.

Q8 What is the role of the activation function in a neural network? Describe three
commonly used activation [Link] is meant by the learning rate? What can
happen if it is too high or too low?

Q9 What do you mean by Curse of dimensionality? Explain any two feature selection
methods.

Q 10 Consider the following neural network with initial weights, biases, and training input
/ outputs as mentioned. Use Sigmoid as the activation function for the hidden as well
as the output layers.
a)​ Given the inputs i1 = 0.05, i2 = 0.10, determine the values of output nodes o1
and o2.
b)​ Calculate the prediction error Etotal if the actual output values o1 and o2 are
0.01 and 0.99 respectively.
c)​ Update the weights b/w hidden layer and output layer using backpropagation
method

You might also like