Cochin University of Science and Techology (CUST) 2007-1st Sem B.Tech Computer Science and Engineering V1ester ,,CS 705(c) ARTIFICIAL NEURAL NETWORK - Question Paper

B.Tech Degree V11 Semester Examination, November 2007
CS 705(c) ARTIFICIAL NEURAL NETWORK

Time: three Hours Maximum Marks: 100

1. a) Explain the biological model of a neuron (6)
b) discuss the exclusive OR-problem (14)
OR
2. a) Differentiate supervised and unsupervised training. (10)
b) What is an activation function? discuss any 2 activation
function (10)

3. a) Explain the training of back propagation network (12)
b) Write notes on network paralysis and local minima (8)
OR
4. a) With the help of neat diagram discuss the architecture of
back propagation network (10)
b) Explain back propagation algorithm with momentum. How
it improves training? (10)

5. a) Explain the architecture of a counter propagation network
with a neat diagram. (12)
b) discuss any 1 application of counter propagation network (8)

OR
6. Explain the training of Kohenen layer (20)

7. a) Explain Bolztmann training (12)
b) discuss simulated annealing process (8)
OR
8. a) Explain the artificial specific heats (10)
b) discuss the issues associated with back propagation network (10)

9. a) Describe the architecture of ART with suitable diagrams (14)
b) discuss any 1 application of Hopfield network (6)
OR
10. a) Explain the structure of BAM with neat diagram (10)
b) discuss how data is stored and retrieved in BAM (10)

Earning:   Approval pending.