Neural Computing(8th Semester B.Tech Kerala University Questions)

NEURAL COMPUTING

MODULE I (4 marks)

1. Write a note on activation function?

2. What is meant by linear separability?

3. Explain the structure of artificial neuron?

4. Define learning and training, in connection with neural networks?

5. Explain unsupervised learning. Give an example?

6. Briefly describe a biological neuron?

7. Define training. What are the major types of neural network learning rules?

8. What is the need for associating a weight or connection strength in the processing element of an artificial network?

9. Explain multilayer feed forward network?

10. What is delta rule?

MODULE II(4 marks)

11. What is neighbourhood as applied to kohonen’s feature maps?

12. Explain the energy landscape in Hopfield net?

13. Explain the behavior of a Hopfield net?

14. Explain the basic concept of Boltzmann machine?

15. Explain adaptive response theory?

16. What is Boltzmann machine?

17. Explain the energy function used in Hopfield networks?

18. Write a note on recurrent auto associative network?

19. How is ART different from other competitive schemes?

20. What is the role of a topological neighbourhood function in kohonen algorithm?

MODULE I (essay: 20 marks)

1. Sketch the basic non-linear model of an artificial neuron .derive the expression of its output. What is meant by activation function?

2. What is X-OR problem and how can you solve it?

3. Explain the back propagation algorithm used in multilayer perceptrons with 3 layers if nodes?

4. Explain the perceptron learning algorithm?

5. What are the limitations of perceptron?

6. What is phonetic typewriter? Explain its operation with a schematic diagram?

7. Explain the architecture of Boltzmann machine. Explain the learning algorithm in detail?

8. Explain multilayer perceptron?

9. Explain counter propagation network?

10. Explain back propagation algorithm and derive the learning rules?

MODULE II(essay:20 marks)

1. Explain Boltzmann machine learning?

2. Explain learning vector quantization?

3. Describe the architecture, training and operation of the discrete Hopfield network?

4. Explain kohonen self organizing networks?

MODULE III (essay: 20 marks)

1. Explain art-2 network?

2. What are the advantages and disadvantages of memory networks?

3. Explain bidirectional associative memory (BAM)?

4. Explain any two applications of neural networks?