Cochin University of Science and Techology (CUST) 2007 B.Tech Electronics and Communications Engineering Artificial Neural Network - Question Paper

Artificial Neural Network (November 2007,old scheme 2000-2001 admission)

1a. discuss exclusive OR issue. How can it be overcome?
b. explain how a perceptron can classify the inputs by linear sperability with suitable examples
2a. describe and discuss the generalized delta rule. How is it various from the original form of the delta rule?
b. explain the various classes of training of artificial neural networks. discuss Hebb’s training algorithm.
3a. discuss back propagation network architecture.
b. discuss the outer product rule
c. write short notes on storage capacity
4a. define in detail the step involved in the back propagation training algorithm
5a. discuss the training in Kohonen’s network
b. why and how processing of input vectors are done in Kohonen’s network
6a. discuss the training of Grossberg layer.
b. discuss 1 typical application of counter propagation network
7a. discuss stimulated annealing in detail.
b. writ short notes on:
i. Boltzman training
ii. Cauchy training
8a. elaborate the issues with back propagation and Cauchy training algorithm?
b. discuss the combined training method
9a. discuss the architecture of ART model
b. elaborate the limitations of Hopfield model?
10a. discuss BAM architecture
b. discuss mutation and cross over in genetic algorithm . elaborate the applications of genetic algorithm?

Earning:   Approval pending.