MORE IN Neural Networks and Fuzzy Systems
MU Electronics Engineering (Semester 8)
Neural Networks and Fuzzy Systems
May 2014
Total marks: --
Total time: --
INSTRUCTIONS
(1) Assume appropriate data and state your reasons
(2) Marks are given to the right of every question
(3) Draw neat diagrams wherever necessary

1 (a) Explain different types of activation functions
5 M
1 (b) Explain k.means of algorithm
5 M
1 (c) Explain any two types of Defuzzification techniques
5 M
1 (d) How many hidden layers are necessary to approximate a continuous function.
5 M

2 (a) Write an algorithm for back propagation training and explain about updating of weight
10 M
2 (b) Explain Hopfield networks in detail.
10 M

3 (a) Using perceptron learning rule, find the. weights required to perform following classifications. Vector (1 1 1 1) and (-1 1-1-1) are the members of first class. Vectors (1 1 1-1) and (1 -1 -1 1) are the member of second class. Use two output neurons. Assume learning rate parameter as 0.9 and initial weight of 0.25. Using training vectors, test the response of net.
10 M
3 (b) What is meant by simulated annealing. Explain procedure of Boltzman machine with its training phase.
10 M

4 (a) Explain the method of solving EX-OR problem using RBF and MLP.
10 M
4 (b) Compare supervised learning with unsupervised learning, Explain with suitable examples.
10 M

5 (a) Explain the operation of fuzzy logic control with process inference block.
10 M
5 (b) Write the properties of fuzzy set theory and explain in detail.
10 M

6 (a) The fuzzy sets are given as follows
$P=\left \{ \frac{0.1}{2}+\frac{0.3}{4}+\frac{0.7}{6}+\frac{0.4}{8} +\frac{0.2}{10}\right \}$
$Q=\left \{ \frac{0.1}{0.1}+\frac{0.3}{0.2}+\frac{0.3}{0.3}+\frac{0.4}{0.4}+\frac{0.5}{0.5}+\frac{0.2}{0.6} \right \}$
$R=\left \{ \frac{0.1}{0}+\frac{0.7}{0.5}+\frac{0.3}{1} \right \$}
Perform the following operations over the fuzzy sets
(i) Max-min composition
(ii) 7 Max product
(iii) Two corss product
10 M
6 (b) Explain Kohonen's self organizing learning algorithm.
10 M

Write a short note on: (any four)
7 (a) Brain state-in-a-box model
5 M
7 (b) Fuzzification Methods
5 M
7 (c) LMS algorithm
5 M
7 (d) Neurodynamic Model
5 M
7 (e) Steepest descent algorithm.
5 M

More question papers from Neural Networks and Fuzzy Systems