Scribd
Upload
24 views16 pages

ML (PCCS-114) Pyq

This document is a question paper for a Machine Learning exam for B.Tech students, consisting of 9 questions across 2 pages. Students are required to attempt any six questions, with each question carrying 10 marks. The paper covers various topics including algorithms, dimensionality reduction, learning systems, and genetic algorithms.

Uploaded by

sainilovepreet178
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
24 views16 pages

ML (PCCS-114) Pyq

This document is a question paper for a Machine Learning exam for B.Tech students, consisting of 9 questions across 2 pages. Students are required to attempt any six questions, with each question carrying 10 marks. The paper covers various topics including algorithms, dimensionality reduction, learning systems, and genetic algorithms.

Uploaded by

sainilovepreet178
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
You are on page 1/ 16

Please check that this question paper contains 9 questions and 2 printed pages within first ten minutes.

[Total No. of Questions: 09] [Total No. of Pages: 2]


Uni. Roll No. …………….
Program: B.Tech.
Semester: 6th
Name of Subject: Machine Learning
Subject Code: PCCS-114
Paper ID: 17190
Time Allowed: 02 Hours Max. Marks: 60

NOTE:

1) Each question is of 10 marks. 13-07-21(E)


2) Attempt any six questions out of nine
3) Any missing data may be assumed appropriately

Q1. Compare Find-S and Candidate Elimination algorithms. Justify the limitations of Find-
S algorithm by taking an example using both the above said algorithms.
Q2. Solve using k-NN for x(A=3, B=7), K=3
A B Label
7 7 False
7 4 False
3 4 True
1 4 True

Q3. Demonstrate the use of Principal Component Analysis for dimensionality reduction
using an example.
Q4. i) ‘A learning system is designed in number of stages.’ Justify this statement by
explaining the stages involved.
ii) Compare supervised learning and unsupervised learning.
Q5. Give valid reasons in favour of the statement ‘Information Gain’ is a good quantitative
measure of the worth of an attribute.’ Also estimate Gain(S,A) and Gain(S,B) for the
given training data
A B Label
a1 b1 No

Page 1 of 2
P.T.O.
a1 b2 Yes
a2 b3 Yes
a2 b2 No
a2 b1 Yes

Q6. i) Distinguish between perceptron rule and delta rule.


ii) Explain how back propagation algorithm works for multilayer feed forward network.
Q7. i) Derive the equation for ‘Brute force learning algorithm’ using Bayes theorem.
ii) Illustrate with an example the significance of a Bayesian belief network.
Q8. Solve the following for genetic algorithms: For the strings of the form x=abcdefgh,
consider the strings x1=65413532, x2=87126601, x3=23921285, x4=41852094. Let
the fitness function be
f(x)= (a+b)-(c+d)+(e+f)-(g+h)
a) Evaluate the fitness of each individual x1,x2,x3,x4
b) Cross the fittest two individuals using one-point crossover at the middle point, with
new hypothesis h1 and h2
c) Cross x1 and x3 using two-point crossover (over points b and f), with new
hypothesis h3 and h4
d) Cross x2 and x3 using a uniform crossover, with new hypothesis h5 and h6
e) Evaluate the fitness of new population, i.e. h1,h2,h3,h4,h5 and h6.
Q9. i) Explain the term ‘Design of Experiments’ with respect to Machine Learning.
ii) Recommend and explain the statistical techniques for estimating quantities about a
population by averaging estimates from multiple small data samples.

***********

Page 2 of 2

You might also like