1
Seat No.: Enrolment
GUJARAT TECHNOLOGICAL UNIVERSITY
ME SEMESTER-1 EXAMINATION WINTER 2018
Subject Code: 3715414 Date: 04/01/2019
Subject Name: Machine Learning
Time: 02:30 PM To 05:00 PM Total Marks: 70
Instructions:
1. Attempt all questions.
2. Make suitable assumptions wherever necessary.
3. Figures to the right indicate full mark.
Q.1

Explain Supervised learning, Unsupervised learning and Reinforcement learning.
07

Define the terms in brief: Bernoulli distribution Binomial distribution Gaussian distribution
07
Q.2

The sales of a company (in million dollars) for each year are shown in the table below
X (Years)
2005
2006
2007
2008
2009
Y (Sales)
12
19
29
37
45
Find the least square regression line .
Use the least square regression line as a model to estimate the sales of the company in 2012.
07

What is machine learning? Give overview of machine learning with suitable example.
07
OR

Consider given dataset in table and using Navie Bayesian classification, find class label for the new dataset
Where



Class Target
H
T
P
Yes
H
F
N
Yes
L
T
P
No
H
T
N
No
M
F
P
No
M
T
N
Yes
L
T
N
Yes
07
Q.3

Is XOR problem solvable using a single perceptron? Justify your answer with appropriate reasoning.
07

Justify: Support vector machine is a good classifier tool.
07
OR
Q.3

Justify "Multilayer perceptron is a must for classifying patterns which are not linearly separable".
07

Write a note on two-category and multi-category case linear discriminant functions.
07
2
Q.4

Explain classification using Back Propagation algorithm with a suitable example.
07

Consider the following data set consisting the height and weight of company workers.
Height
185
170
168
179
182
188
180
183
Weight
72
56
60
68
72
77
71
84
Use algorithm with k=2 and Euclidian distance as the distance measure. Assume and as the initial cluster centroids.
07
OR
Q.4

Write a short note on Deep learning.
07

Take 10 points in two dimensions having coordinate values as: Use k-means clustering to cluster them into two clusters. Assume and as the initial cluster centroids. Show computation up to two iterations and report the centroids after two iterations. Assume Euclidean distance as the distance measure.
07
Q.5

Define the term over fitting. What are the possible reasons for over fitting?
07

Explain algorithm. Is the algorithm sensitive to choice of initial centroid? Justify your answer.
07
OR
Q.5

What is Occam's Razor?
07

List and explain application of Machine Learning algorithms.
07