Total No. of Questions : 10] SEAT No.
8
23
P976 [Total No. of Pages : 3
ic-
tat
[5871]-753
9s
B.E. (Honours)
4:2
02 91
ARTIFICIAL INTELLIGENCE AND MACHINE
3:4
Soft Computing and Deep Learning
0
21
7/0 13
(2015 Pattern) (Semester - II) (410303)
0
Time : 2½ Hours] 7/2 [Max. Marks : 70
.23 GP
Instructions to the candidates:
1) Answer Q.1 or Q.2, Q.3 or Q.4, Q.5 or Q.6, Q.7 or Q.8. Q. 9 or Q. 10.
E
2) Assume suitable data, if necessary.
80
8
C
23
3) Neat diagrams must be drawn whenever necessary.
ic-
4) Figures to the right indicate full marks.
16
tat
8.2
9s
Q1) a) Differentiate Supervised Learning and Unsupervised Learning. [3]
.24
4:2
91
b) What is soft computing? Indicate biological analogies of basic techniques
49
3:4
of soft computing. Describe why soft computing is particularly useful in
30
21
representing and reasoning with human-oriented knowledge. [6]
01
02
7/2
OR
GP
7/0
Q2) a) Define the following for membership functions: [3]
CE
80
8
23
i) Core
.23
ic-
16
tat
ii) Support
8.2
9s
.24
4:2
iii) Boundary
91
49
3:4
b) Define Crisp and Fuzzy relation. Describe fuzzy max-min composition,
30
21
fuzzy min-max composition and max-product composition. [6]
01
02
7/2
GP
7/0
Q3) a) Explain in detail Mamdani FIS.(Fuzzy Inference System) [3]
CE
80
b) Explain fuzzification and defuzzification techniques. [6]
.23
16
OR
8.2
.24
P.T.O.
49
Q4) a) Explain in detail convergence working principle in genetic algorithm. [3]
8
23
b) What is genetic programming? Write and explain the executional steps
ic-
of Genetic Programming. [6]
tat
9s
4:2
02 91
Q5) a) Explain working of Bayesian Neural Networks with suitable diagram. [6]
3:4
0
b) What is Neural Network? Explain with suitable diagram the different
21
7/0 13
classes of network architecture. [6]
0
7/2
c) Explain the following terms: [4]
.23 GP
i) Theano in Deep Learning.
E
80
8
C
23
ii) Caffe in Deep Learning.
ic-
OR
16
tat
8.2
9s
Q6) a) Explain in detail various Open Source Deep Learning Libraries and
.24
4:2
Platforms. [6]
91
49
3:4
b) What do you mean by Tensor in TensorFlow? What are the Programming
30
21
elements of TensorFlow? [6]
01
02
c) Explain in detail Feed Forward Neural Network. [4]
7/2
GP
7/0
CE
80
Q7) a) Explain with example Self Organizing Maps. [8]
8
23
.23
ic-
b) What is a Deep Belief Network? How did Deep Belief Neural Networks
16
Evolve? [6]
tat
8.2
9s
c) What is image captioning in Deep Learning? [4]
.24
4:2
91
OR
49
3:4
30
21
Q8) a) What is the role of the Activation functions in Neural Networks? Explain
in detail some of the popular Activation Functions used in Neural
01
02
Networks. [8]
7/2
GP
7/0
b) What is pooling? What are the different types of Pooling? Explain their
CE
characteristics. [6]
80
.23
c) What do you mean by Perceptron? Explain in detail different types of
16
Perceptrons? [4]
8.2
.24
[5871]-753 2
49
Q9) a) Explain following with example [8]
8
23
i) Sequential processing LSTM
ic-
tat
ii) Recurrent Neural Network(RNN)
9s
4:2
b) Explain in detail architecture Convolution Neural Network (CNN). Write
02 91
down various properties of CNN representations. [6]
3:4
0
21
c) Explain on detail Architecture of Back Propagation (BP) Networks. [4]
7/0 13
0
OR
7/2
.23 GP
Q10)a) What is associate memory in neural network? Explain various types and
E
80
applications of associative memory. [8]
8
C
23
ic-
b) Write short notes on: [6]
16
tat
i) Artificial Neural Network (ANN)
8.2
9s
.24
4:2
ii) Application of RNN
91
49
3:4
c) Explain Adaptive Resonance Theory? What is the major difference
30
21
between ART1 and ART2? [4]
01
02
7/2
GP
7/0
CE
80
8
23
.23
ic-
16
tat
8.2
9s
.24
4:2
91
49
3:4
30
21
01
02
7/2
GP
7/0
CE
80
.23
16
8.2
.24
[5871]-753 3
49
