Total No. of Questions:8| SEAT No.

:TI900S8 709
P-7619
Total No. of P'ages :3

|6180|-139
T.E. (Information Technology)
MÁCHINE LEARNING
(2019 Pattern) (Semester - 1I) (314443)
Time :27: Hours] /Mar. Marks : 70
Instructions to the c¡ndidates .:
I) Ansswer Qlor Q2. Q3 or Q4. Q5 or Q6, Q7 or Q8.
2) Ncat iagrans must be drawn wherever necessary.
3) Figures to the right side indicate full marks.
4) Assume Suitable data if necessary.

Q1) a) Compare univariate and multivariate linear regression. (CO3,L2] (6|

b Describe the tradeoff between bias and variance using dart example.
|CO03,LI| [5|
c) Compute the R-Square error of the best fit line Y =- 0.94X+43.7" for
below input instance space |CO3, L3| |7]
Sr.No. X

3 40
2 10 35

30

4 15 32

22 19

6 22 26

7 23 24

8 28

28 18 eteua
10 35

OR
PLO.
 02) a) Explain gradient descent technique for optimization in linear regression
with example. |C03, LI|
b) Explain the cost function usedto evaluate the performance ofregression.
|CO3, L3] |6|
c) What is least square method?Explain least square method in the context
of regression. |CO3, L!| |6|

Q3) a) Describe Bayesian network in short for learning and inferences.

|C04, L3 |8]
b) Explajn naïv bays algorithm? For below dataset compute the posterior
probab+lity of banana class using Bayes rule. [C04, L3| I71
Fruit Yellow Sweet Total
Long
Orange 350 450 650
Banana 400 300 350 400
Other 50 100 50 50
Total 800 850 400 1200
c) Write any four applications of naive Báyès classifier. [C04, L1| |2]
OR
04) a) Explain ID-3 decision tree algorithm in detail with
66.9
example.(CO4, L2||8|
b) Explain the following measures of impurity with
example.[C04, L2||9]
i) Information Gain Gini Index i) Entropy

Q5) a) Explain K-Nearest Neighbor algorithm with example. |CO5, L2] 17]
b) Suppose we have the following dataset that has
various transactions and
from this dataset, we need to find the frequent item sets ànd
association rules using the Apriorialgorithm. Given minimumgenerate the
support =2
and minimum confidence = 50%. |CO5, L4]
|8|
TID ITEM SET TID ITEM SET
TI A,B T6 B,C
T2 B,D T7 AC
T3 B.C T8 A, BC,E
T4 A, B, D T9 A, B. C
T5 A. C

|6180]-139 2
sion c) Explain any one of the following ternns: |C05, L2| 31

i) Medoid OR 1) Dendrogram
OR
locations) into
V6) a) Cluster the following eight points (with (x, y) representing
clusters: A 1(2, 10), A2(2, 5), A3(8, 4), A4(5, 8), A5(7, 5), A6(6, 4).
three Al(2,10), A4(5, 8) and A7(1.
A7(1,2), A8(4,9) Initial cluster centers are:
a = (xl, yl) andb=(x2, y2)
2). The distance function between two points Algorithmto find the
is defined as-P(a, b)= (x2-xI|+|y2yl| Use K-Means
iteration. [C05, L3] 19|
three cluster centers after the first
following performance
b) Explain the Apriori algorithm in brief. Explain the |6|
measure of association rule mining. |CO5, L2]
1)Support 2) Confidence 3) Lift
example. (C05, LI|
Explain any one of the following distance metrics with |3|
Euclidean Distance 2) Manhattan Distance
1)
3) Hamming Distance
Q) a) Explain perceptron learning algorithm? Describe shortly about how learning
parametersare updated in multi-layer perceptron. (CO6, L3] |8]
b) Explain Sigmoid, Tanh and Relu activation functions in detail. |CO6. L2|
[9|
OR
98) a) Explain the simulation ofANDgate using McCulloch Pitts Neuron? What
is vanishing gradient desçent problem. |CO6, L2| 9

b) Explain what is Deep Learning and its different architectures? State the
[CO6, L2]various applications of deep learning. |8|

|61801-139