B.E-B.E Computer Engineering OtherNEURAL NETWORKS (Karunya University, Coimbatore-2012)
Saturday, 24 August 2013 12:59anudouglas
Karunya University
(Karunya Institute of Technology and Sciences)
(Declared as Deemed to be University under Sec.3 of the UGC Act, 1956)
End Semester Examination – November/ December - 2012
Subject Title: NEURAL NETWORKS Time: 3 hours
Subject Code: 09EC202 Maximum Marks: 100
Answer ALL questions
PART – A (10 x 1 = 10 MARKS)1. List out the differences between artificial neural network and Biological network.
2. What are Dendrites?
3. What is meant by connectivity?
4. Define Bias.
5. Mention the limitations of Hebb Net.
6. What is meant by discrete and continuous perceptron models?
7. What is meant by multi-layer feed forward networks?
8. Define Learning Rate parameter.
9. What is Hebbian Learning?
10. What is auto associative memory?
PART – B (5 x 3 = 15 MARKS)
11. State the differences between Humans and computers.
12. Explain about classification taxonomy of ANN.
13. What is learning in Neural Networks?
14. Explain about alternative activation function in Multilayer Perceptron (MLP).
15. Discuss about the energy function used in the discrete Hopfield networks.
PART – C (5 x 15 = 75 MARKS)
16. a. With neat diagrams explain the organization of a Human brain. (7)
b. Describe McCulloch-Pitts model with diagrams. (8)
(OR)
17. a. Summarize elaborately about the history of Neural Networks in terms of architecture and algorithm developments. (9)
b. Explain the characteristics and applications of ANN. (6)
18. a. Explain about different activation functions in detail. (7)
b. Explain in detail about the architecture of Artificial Neural Networks with operations.
(8)
(OR)
19. a. Explain in detail about three different learning strategies. (7)
b. Explain the concepts behind the different learning rules used in ANN. (8)
20. a. Discuss the limitations in the Perceptron model. (4)
b. Give the Perceptron model to solve two input AND, OR, Exclusive OR problems and explain. (11)
(OR)
[P.T.O]
21. a. Draw the architecture of a Single layer perceptron (SLP) and explain its operation. Mention its advantages and disadvantages. (10)
b. Write the Perceptron training algorithm for several output classes. (5)
22. Explain with a neat diagram about the multi-layer back propagation network and derive the generalized delta rule.
(OR)
23. Explain in detail about the different practical issues in the BPNN and give the remedy to overcome it.
24. With network architecture, explain the Continuous Hopfield Network.
(OR)
25. Draw and explain the BAM architecture and its storage and recall training algorithms.
| Earning: ₹ 5.50/- |
