Madras University (UnOM) 2006 M.Sc Information Technology Artificial Neural Network - Question Paper

Time: 3 hours
Maximum: 75 marks

PART A - [5 x five = Marks 25]

ans ALL ques..
All ques. carry equal marks.

1. (a) discuss the neural system hierarchies, with examples.

Or

(b) define self organisation model.

2. (a) discuss correlation based learning method.

Or

(b) What is a constraint based neural network? explain.

3. (a) define the incremental learning principle.

Or

(b) discuss fault measures and fault trajectories for ANN learning approaches.

4. (a) explain about control networks.

Or

(b) discuss hybrid models and compare with parallel models.

5. (a) What is spatio temporal neural network? explain.

Or

(b) elaborate symbolic schemes? discuss them.

PART B - [5 x 10 = Marks 50]

ans any 5 ques..
All ques. carry equal marks.

6. discuss the basic concepts of neural networks.

7. define Back propagation network and discuss how fault is minimized.

8. discuss different learning schemes used in NN.

9. define mathematical modeling for learning procedure.

10. discuss the differentiation models for neural networks.

11. explain the significance of symbolic methods in NN.

12. define knowledge based approaches in learning and compare with other schemes.

13. Write short notes on:
(a) Optimization models
(b) Applications of neural networks.

Earning:   Approval pending.