Total No. of Questions : 8] SEAT No.
8
23
PB-3860 [Total No. of Pages : 4
ic-
[6262]-123
tat
2s
T.E. (IT)
8:1
02 91
MACHINE LEARNING
9:4
0
(2019 Pattern) (Semester - I) (314443)
40
2/0 13
0
5/2
Time : 2½ Hours] [Max. Marks : 70
.23 GP
Instructions to the candidates:
E
1) Answer Q1 or Q2, Q3 or Q4, Q5 or Q6, Q7 or Q8.
82
8
C
23
2) Neat diagrams must be drawn wherever necessary.
ic-
3) Figures to the right side indicate full marks.
16
tat
4) Assume Suitable data if necessary.
8.2
2s
.24
8:1
91
Q1) a) State and explain need of Regression analysis. [4]
49
9:4
b) How gradient descent does helps to optimize linear regression model?[6]
30
40
c) What are the different ways to prevent overfitting and underfitting.
01
02
[8]
5/2
GP
2/0
OR
CE
Q2) a) What are different cost functions to access the performance of linear
82
8
23
Regression model? In the given Dataset the outliers represent anomalies.
.23
ic-
Which cost function will be more suitable and why? [5]
16
tat
b) Define of Multivariate Regression and State advantages and disadvantages
8.2
2s
of Multivariate Regression. [5]
.24
8:1
c) Consider the following data : [8]
91
49
9:4
Prize in Rs Amount Demanded
30
40
1 10 40
01
02
2 11 38
5/2
GP
3 16 48
2/0
4 18 40
CE
82
5 20 60
.23
i) Find values of 0 and 1 w.r.t. linear regression model which best
16
fits given data.
8.2
ii) Interpret and explain equation of regression line.
.24
iii) Estimate the likely demand when the price is Rs.15.
49
P.T.O.
Q3) a) Consider following data. Which feature will be selected as a root node?
8
23
Use Information Gain. Played football is dependent feature. [10]
ic-
Outlook Temperature Humidity Wind Played football (Yes/No)
tat
Sunny Hot High Weak No
2s
Sunny Hot High Strong No
8:1
Overcast Hot High Weak Yes
02 91
9:4
Rain Mild High Weak Yes
0
40
Rain Cool Normal Weak Yes
Rain
2/0 13
Cool Normal Strong No
0
5/2
Overcast Cool Normal Strong Yes
.23 GP
Sunny Mild High Weak No
Sunny Cool Normal Weak Yes
E
82
8
C
23
Rain Mild Normal Weak Yes
ic-
Sunny Mild Normal Strong Yes
16
tat
Overcast Mild High Strong Yes
8.2
2s
Overcast Hot Normal Weak Yes
.24
8:1
Rain Mild High Strong No
91
49
9:4
b) Define and Explain following terms [7]
30
i) Bayesian Network
40
ii) Advantages and disadvantages of Naïve Bayes Classifier
01
02
OR
5/2
GP
Q4) a) For the given data set apply Naive Bayes Classifier and predict the Class
2/0
for Type of Type of family structure=Single Parent, Age group=Young,
CE
82
8
and Income status=Low [10]
23
.23
ic-
Type of family Age group Income status Will they buy a car
16
tat
structure
8.2
2s
Nuclear Young Low Yes
.24
8:1
91
Extended Old Low No
49
9:4
30
Childless Middle-aged Low No
40
Childless Young Medium Yes
01
02
Single Parent Middle-aged Medium Yes
5/2
GP
Childless Young Low No
2/0
CE
Nuclear Old High Yes
82
.23
Nuclear Middle-aged Medium Yes
16
Extended Middle-aged High Yes
8.2
Single Parent Old Low No
.24
49
[6262]-123 2
b) Define and explain following terms [7]
8
23
i) Minority Class
ic-
ii) Gini Index
tat
iii) Entropy
2s
8:1
iv) Information Gain
02 91
9:4
0
40
Q5) a) Find all association rules in the following database in the following database
2/0 13
with minimum support =2 and minimum confidence = 75%. [10]
0
5/2
Transactions Data Items
.23 GP
1 Bread, Milk, Diaper
E
82
8
2 Bread, Milk, Diaper, Coke
C
23
3 Diaper, Beer, Eggs
ic-
16
tat
4 Bread, Milk, Coke
8.2
2s
b) State & explain with appropriate example different types of linkage use
.24
8:1
in clustering. [8]
91
49
9:4
OR
30
40
Q6) a) Explain following Terms [8]
01
02
i) Rule
5/2
GP
ii) Support
2/0
iii) Lift
CE
82
8
iv) Confidence
23
.23
ic-
b) Apply KNN on the Following data. Find class of person whose height is
16
tat
170 cm and weight is 57 kg. Consider value K =5 and use Euclidian distance
8.2
2s
formula. [10]
.24
8:1
Height (CM) Weight (KG) Class
91
49
9:4
167 51 Underweight
30
40
182 62 Normal
01
02
176 69 Normal
5/2
GP
173 64 Normal
2/0
172 65 Normal
CE
82
174 56 Underweight
.23
169 58 Normal
16
173 57 Normal
8.2
170 55 Normal
.24
49
[6262]-123 3
Q7) a) With the help of suitable diagram explain Biological Neuron. [5]
8
23
b) Explain the architecture of feed forward neural network. State its
ic-
limitations. [7]
tat
2s
c) What is deep learning? Explain different applications of deep learning.
8:1
[5]
02 91
9:4
OR
0
40
Q8) a) 2/0 13
What is perceptron? Explain multilayer perceptron in detail. [5]
0
b) Explain why we use non-linearity function? State and explain three types of
5/2
.23 GP
neurons that add non-linearity in their computations. [7]
E
c) What is ANN? Explain McCulloch Pitts Neuron. [5]
82
8
C
23
ic-
16
tat
8.2
2s
.24
8:1
91
49
9:4
30
40
01
02
5/2
GP
2/0
CE
82
8
23
.23
ic-
16
tat
8.2
2s
.24
8:1
91
49
9:4
30
40
01
02
5/2
GP
2/0
CE
82
.23
16
8.2
.24
49
[6262]-123 4
