Cochin University of Science and Techology (CUST) 2008 B.Tech Electronics and Communications Engineering Aritficial neural networks - Question Paper

artificial neural networks(supplementary,june2008)


1a discuss the characteristics of artificial neural neuron.
b compare in detail artificial neuron and biological neuron
c differentiate single layer neural network and multilayer neural network
2a discuss with examples supervised and unsupervised training of artificial neural network
b discuss exclusive OR issue of preceptron and how you overcome that
3a discuss in detail the architecture of Back propagation network
b discuss how training happens in different layers of Back propagation network
4 write short notes on:
i. network analysis
ii. local minima
iii. temporal unstability
5 discuss the architecture of counter propagation network in detail with 1 application as example
6 discuss the significance of Kohonen layer and Grossberg layer in counter propagation network
7 write short notes on:
i. simulated annealing
ii. Boltzmann training
iii. Cauchy training
iv. Artificial specific heat method
8a. discuss in detail non linear optimization issue and how it can be solved using artificial neural network
b. discuss how network paralysis is overcome with combined method of training
9a. discuss in detail architecture of ART
b. discuss what is statistical Hopfield network citing with any 1 application
10a discuss mutation and cross over operations in genetic algorithm
b. discuss how encoding and retrieving of association happens in bidirectional associative memories.

Earning:   Approval pending.