Pune University Soft Computing Exam Papers

Pune University Soft Computing Exam Papers

B.E. (E & TC) SOFT COMPUTING (2008 Pattern) (Sem. – II) (Elective – III)

Time :3 Hours]

Instructions to the candidates:-

1)             Question Nos. 1 and 12 are compulsory. Out of the remaining attempt 2 questions from Section I and 2 questions from Section II.

2)            Answers to the two sections should be written in separate books.

3)            Neat diagrams must be drawn wherever necessary.

4)             Pigures to the right indicate full marks.

5)             Use of logarithmic tables, slide rule, Mollier charts, electronic pocket calculator and steam tables is allowed.

6)            Assume suitable data, if necessary.

SECTION – I

QI) Write notes on (any three)                                                                                          [18]

a)            Applications of Soft Computing.

b)            Hybrid systems.

c)            Neuro-Fuzzy and Soft Computing characteristics.

d)            Compare and contrast hard and soft computing.

OR

Q2) a) Define a fuzzy set and explain the concept of a fuzzy number. What is the significance of fuzziness.                                                                                                             [8]

b)             Consider two fuzzy sets A and B find Complement, Union, Intersection, Difference, and De Morgan’s laws:                                                                                     [8]

A = |0.8 0.4 0.6 0.1 0.31

=1 2 ’ 3 ’ 4 ’ 5 ’ 6 J

B = 10.3 0.8 0.6 0.8 021

=1 2 ’ 3 ’ 4 ’ 5 ’ 6 J

 

Explain any four fuzzy membership functions with their transfer characteristics.                                                                                                                                         [8]

f0.2 0.5 0.71

Given a rule : IF x is A, THEN y is B, where A – j 1 ’ 2 ’ 3 J and

B_f06 0.8 0.41

_l 5 ’ 7 ’ 9 J•

Infer B’ for another rule: IF x is A’, THEN y is B’, where f0.5 0.9 0.31

A’ = j 1 ’ 2 ’ 3 J, using Mamdani Implication rule and max-min composition.                                                                                                                                      [8]

OR

Describe the architecture of a Mamdani type Fuzzy Logic Controller and compare it with a conventional PID controller.                                                                              [8]

What are the principal design parameters of a Fuzzy Logic Controller? Explain with a suitable example.                                                                                                     [8]

What are the advantages of Fuzzy Logic Controller over that of a conventional controller.                                                                                                                          [8]

Explain the Sugeno Fuzzy Inference Model with a suitable example. [8]

OR

Define the following terms with reference to fuzzy inference systems: [6]

i)         Premise (Antecedent)

ii)        Conclusion (Consequent)

iii)       Rule – base.

Given two rules:                                                                                                                                                                                                                                [IO]

RULE I : if height is “TALL”, then speed is “HIGH”

RULE 2 : if height is “MEDIUM”, then speed is “MODERATE”

The fuzzy sets for height (in feet) and speed (in m/s) are:

f0.5 0.8 11                                                                                                   f0.4 0.7 0.91

H = “TALL” = j — ’—’- \, S = “HIGH” =

5 6 7 J 1                                I 579 J

f0.6 0.7 0.61

H 2 = “MEDIUM” = j^’^’^ J, S2 = “MODERATE” f0.6 0.8 0.71

 

f0.5 0.9 0.81

For a given H’ = “ABOVE AVERAGE” = j — ,—, — J,

Compute S’ = “ABOVE NORMAL”

SECTION – II

Q7) a) State the various learning rules in neural networks.                         [8]

b)            Using Mc-Culloch Pitts neuron, implement a bipolar AND function. Assume initial weights to be [I I].                                                                                              [8]

OR

Q8) a) What is a perceptron network? State the algorithm for perceptron learning.

[8]

b)            Train a perceptron network for learning a binary OR gate function. Work out two complete iterations.                                                                                            [8]

Q9) a) Explain backpropagation algorithm for MLP with a neat signal flow graph. [8]

b)            Enlist the various activations functions used in neural networks and explain any two in details.                                                                                                                  [8]

OR

QIO) State the applications of artificial neural networks and explain any two in details.   [IT]

QII) a) Explain unsupervised learning mechanism in contrast with a supervised learning mechanism.                                                                                                         [8]

b)            Describe the Self Organizing Map architecture and explain the Kohonen model.       [8]

OR

Q12)Write notes on (any two)                                                      [18]

a)            Architecture of ANFIS

b)            Advantages of ANFIS over FIS

c)            Use of ANN in process control.

Leave a Comment